I. BANKRUPTCY AUCTIONS-LEGAL AND ECONOMIC PERSPECTIVES 330
A. Legal Background 330
1. Î¤ypical Scenarios 330
2. The Question of Credit Bidding 333
B. Economic Environment 336
1. Why an Auction? 336
2. Challenges for Bankruptcy Auctions 345
II. AUCTION THEORY-A PRIMER 352
A. General 353
?. A Typology of Auctions 354
C. Choosing an Auction Method 358
1. Sealed-Bid Auction and Dutch Auction 358
2. English Auction and Second-Price Auction 359
3. Why Are Second-Price Auctions So Rare? 361
4. Sealed-Bid Auction Versus English Auction 363
a. When Is the English Auction Preferable?.. 363
1. Transaction Costs and Erroneous Strategy 363
2. Incomplete Information and the "Winner's Curse" 364
b. When Is the Sealed-Bid Auction Preferable?.. 366
D. Additional Aspects 370
1. Information on the Asset 370
2. Reserve Price 371
E. Auction Theory in Action-The FCC Spectrum Licenses Auctions 373
F. The Lesson Learned 379
III. DESIGNING AN OPTIMAL BANKRUPTCY AUCTION 380
A. General Guidelines 381
1. Utilizing Auction Theory 381
2. The Suggested ADVA Design 382
a. The Anglo-Dutch Feature 384
b. The Veto Feature 385
c. Other Features 387
B. Credit Bidding-A Critical Analysis 388
C. The Absence of a Regulator 391
IV. CONCLUSION 393
Whether in a bankruptcy case of a fraudulently operated firm such as Enron1 or that of the temporarily distressed Los Angeles Dodgers baseball team;2 in a routine and small-scale bankruptcy case of a neighborhood restaurant3 or in the unique multi-billion dollar bailouts of Chrysler and General Motors;4 in a bankruptcy of a traditional manufacturing firm5 or of a high-tech business;6 in the United States or in Europe;7 bankruptcy auctions have become "no longer a last resort but an option to be exercised at any time if it is in the creditors' interest."8 Indeed, recent studies show that over half of the firms entering corporate bankruptcy proceedings9 are auctioned10-either piecemeal or as a going concern-instead of reorganized" and restructured.12 Auctions seem to create value for financially distressed firms.13 However, what is the optimal design of a bankruptcy auction? Does the method by which a bankruptcy auction is executed matter at all? Is there a difference, for example, between the impact of an "English auction" and that of a "sealedbid" auction? Does it matter whether a reserve price is set in advance or whether that price is hidden from potential bidders? More generally, does one auction method fit all bankrupt firms?
Surprisingly, although bankruptcy auctions are ubiquitous, these questions have not yet been answered, as auction design in the specific context of bankruptcy has barely been studied.14 Thus, the optimal design of a bankruptcy auction remains a "black box" to lawmakers and practitioners.15 Bankruptcy scholars have argued vigorously about whether bankruptcy auctions should be employed as the preferred method for redeploying the bankrupt firm's assets.16 Additionally, they have also utilized empirical data gathered on bankruptcy auctions worldwide to argue over the relative efficacy of auctions in bankruptcy cases.17 However, little attention has been given to the art of designing an optimal bankruptcy auction. Even more importantly, much of the work performed by scientists in the context of general auctions-many of the valuable insights generated and findings gathered during decades of research-has failed to reach and influence the setting of bankruptcy auctions.
This Article contributes to this scarce literature by shedding light on the possible interaction of bankruptcy auctions with a field of study that has strongly influenced the design of auctions in general: game theory. Our main theme concerns the need to consider the insights offered by the field of "auction theory" when planning the manner in which the assets of financially distressed firms are to be auctioned within a bankruptcy proceeding.18
However, the contribution of auction theory to the regulation of bankruptcy auction design is nuanced and should be examined using a wide perspective. Indeed, one can speak of both the merits and the procedural contributions of auction theory to the practice of bankruptcy auctions. We illustrate the former by examining the case of "credit bidding."19 Having become the subject of a circuit split and a ruling by the Supreme Court, credit bidding-the practice of allowing a secured creditor to bid in a bankruptcy auction using its claim as currency (in whole or part) rather than cash-can be usefully analyzed with the tools provided by auction theory.
Thus, this Article offers three contributions to the design of auctions in bankruptcy. First, we argue that as a default design (which may be changed under special circumstances), the bankruptcy auctions of large firms should be executed according to a novel design that we call "Anglo-Dutch Veto Auction" ("ADVA").20 An ADVA design consists of an Anglo-Dutch auction-rather than a simple sealed-bid or English auction21-at the end of which the secured creditor is extended the right to veto the sale in exchange for paying a pre-defined amount as the cost to the highest bidder and conditioned upon the bidding price not exceeding the amount of the secured claim. We argue that an ADVA design is superior mainly because it is the best way to simultaneously tackle problems of weak competition and lack of information about the auctioned assets-two problems that are the most critical in the context of bankruptcy auctions.
Second, it is argued that, although "credit bidding" may safeguard the secured creditor, such protection becomes redundant once the auction is executed as an ADVA. Given that "credit bidding" poses a possible danger to the intensity of competition in the auction and (occasionally) to unsecured creditors, the conclusion 99 is that "credit bidding" should not be encouraged."
Third, in the context of a procedural contribution of auction theory to the design of bankruptcy auctions, this Article argues that one of the main problems-which should also alert lawmakers- undermining any effort to execute efficient bankruptcy auctions concerns the absence of a regulator who can implement a structured, trial-and-error procedure. For example, unlike FCC auctions, where the FCC can employ auction theory to design what it believes to be the optimal auction given the circumstances, execute the auction, closely follow the results, and re-design a better auction in the future, bankruptcy auctions currently do not follow a similar track. The task of designing each bankruptcy auction is privatized, and although an accepted best practice may evolve among practitioners,24 this practice can hardly entertain any attempt to improve-whether ad hoc or permanently-the design of these auctions, unless the bankruptcy courts decide to increase their involvement in regulating the specific auction designs.
The Article proceeds as follows. Part I describes the legal and economic landscape of the auctions executed in corporate bankruptcy proceedings. This Part begins with a general account of the legal framework for bankruptcy auctions as engendered by the Bankruptcy Code and then proceeds to discuss the legal debate regarding "credit bidding." Next follows a comprehensive summary of the economic environment of bankruptcy auctions. The reasons why bankruptcy auctions have become so widespread are discussed, as are the problems that undermine these auctions.
Part II is a primer describing the influence of game theory on auctions in general. Specifically, this Part refers to the field of knowledge recognized as auction theory. This Part starts with a general discussion of the goals of auction theory and then proceeds to offer a typology of common auctions. The various considerations relevant to choosing a particular auction design are discussed, as are various other aspects that concern the execution of an auction. To establish the practicality and relevance of auction theory, this Article describes the manner in which auction theory has been uniquely employed by the FCC to design its spectrum license auctions since the 1990s.
Parts I and II set the necessary background for the normative discussion in Part III by providing a comprehensive account of bankruptcy auctions and auction theory, respectively. However, it should be noted that the considerable length of each of these Parts is inevitable, mostly because of the absence of an existing legal text to which one could refer the reader. Nonetheless, a reader educated in either bankruptcy auctions or auction theory can skip the corresponding Part.
Part III focuses on the possible contributions of auction theory to the execution of bankruptcy auctions. This Part begins by discussing the need to consider auction theory when designing a bankruptcy auction and then develops a suggested default design- the ADVA-which we argue should serve as a baseline model of a bankruptcy auction. The question of "credit bidding" is then analyzed, followed by a discussion of the procedural aspects associated with employing auction theory to design optimal bankruptcy auctions. A conclusion follows.
I. Bankruptcy Auctions-Legal and Economic Perspectives
The purpose of this Part is to provide the reader with a succinct yet rich description of the bankruptcy auctions landscape. Section A will describe the legal background against which bankruptcy auctions are executed in the United States. Section ? will discuss the relevant economic context.
A. Legal Background
1. Typical Scenarios
As a bankruptcy proceeding commences,25 a bankruptcy auction may occur under one of two typical scenarios. In the first and more conventional scenario, an auction is conducted as the firm is either liquidated (in a Chapter 7 proceeding26) or reorganized (in a Chapter 11 proceeding27). The Chapter 7 auction is an obvious move because the firm must be shut down and its assets sold.28 However, Chapter 11 auctions are also prevalent. The goal of a Chapter 11 proceeding is to enable the claimants of the bankrupt firm to negotiate a plan among themselves. This plan is then put to a vote and must be accepted by the claimants and confirmed by the bankruptcy court. The motivation for this framework is that the claimants may agree on the manner in which the financially distressed firm is to be both restructured and reorganized. As part of the firm's reorganization, the claimants may agree upon an auction, whose purpose is to convert assets into cash.
The second scenario that typically ends in a bankruptcy auction is a sale under Section 363(b) of the Bankruptcy Code, which states that "[t]he trustee, after notice and a hearing, may use, sell, or lease, other than in the ordinary course of business, property of the estate ... ."29 A Section 363(b) sale is allowed not only for trustees in Chapter 7 liquidations, but also as an out-of-plan maneuver for debtors-in-possession-usually the bankrupt firm's incumbent management-during a Chapter 11 proceeding. 0 In the Chapter 11 context, Section 363(b) presents an anomaly, as it bypasses rather easily the classic and carefully designed Chapter 11 structure of negotiation-plan-confirmation.31
This anomaly is accentuated in those cases in which the debtor-in-possession auctions most, if not all, of the firm's assets. Indeed, Section 363(b) was originally enacted as a "side door" to cope with unique cases of bankrupt firms that may be considered to be "melting ice cubes."33 For these firms, each passing day either deteriorates the firm's finances (e.g., declining cash reserves) or depreciates the value of its assets (e.g., high market values are about to decline).34 For this reason, Section 363(b) entertains a path for a rather short-order sale. The more serious barrier to Section 363(b) sales is the judicially imposed requirement that a business justification be provided for the sale to obtain the court's approval.35 The "business justification" requirement entrusts the bankruptcy court with discretion, and the courts were instructed to consider several factors when contemplating the approval of a 363(b) sale.36 These factors include the proportionate value of the asset to the estate as a whole, the effect of the proposed disposition on future reorganization plans, and, perhaps most importantly, whether the Ï 7 asset is increasing or decreasing in value. The courts were also instructed to consider whether the sale was adequately and reasonably noticed, whether it was proposed in good faith, and whether the disposition of the assets is "fair and expeditious."38
The extent to which Section 363(b) sales have been employed to dispose of the bankrupt firm, even in Chapter 11, has increased significantly over the years.39 During the 1980s, bankruptcy courts were reluctant to allow such sales to proceed. However, during the 1990s, the courts gradually released the harness, and the flow of such sales has only increased since.40
2. The Question of Credit Bidding
Auctioning the assets of financially distressed firms undergoing a bankruptcy proceeding may evoke several legal questions. One such question that is currently preoccupying the courts is the issue of credit bidding. The legal question is fairly simple: can secured creditors participate in a bankruptcy auction bidding "with their claim" rather than with cash? In other words, if the bankrupt firm has a secured creditor and its encumbered assets are auctioned, can that secured creditor-when she chooses to compete over the auctioned asset-submit bids that offer a set-off of her claim for consideration? Obviously, the secured creditor can bid in this way up to the face value of her secured claim. However, is she entitled to do so? Specifically, if the debtor is not interested in the secured creditor bidding in this manner, can the debtor prevent the secured creditor from doing so?41
Because bankruptcy auctions can be held as a Section 363(b) sale or follow a confirmed reorganization plan, the question regarding credit bidding is actually more concrete. In the context of a Section 363(b) sale, there currently seems to be no legal problem, as Section 363(k) specifically states that the secured creditor is entitled to credit bid. 2 According to this provision, but "for cause," the secured creditor cannot be prevented from credit bidding in a Section 363(b) sale.43
However, things become more complicated in the context of a confirmed reorganization plan sale. Generally, for the bankruptcy court to confirm a reorganization plan, the debtor must prove, inter alia, that either each class of claims or interests has accepted the plan or the plan does not impair that class's claims or interests.44 In a typical credit-bidding conflict under a reorganization plan, the interests of the secured creditor are impaired because, according to the plan, the encumbered assets are usually to be sold without the liens securing the claims surviving the auction and attaching to the assets as they are transferred to the buyer.45 Denied the possibility of credit bidding, the secured creditor opposes the reorganization plan and does not accept it.46 This rejection causes the debtor interested in confirming the plan to seek a "cram down." Indeed, the Bankruptcy Code entertains the option of confirming a "cram down" plan, which is an impairing plan that has not been accepted by all classes of claims and is therefore "crammed down the throats of objecting creditors."47 According to the Code, the court may confirm an impairing plan that has not been accepted by all classes if the plan "does not discriminate unfairly, and is fair and equitable" with respect to each impaired, non-accepting class.48 For a plan to be "fair and equitable" with respect to a class of secured claims, the court must be convinced that the plan provides the following:
(ii) for the sale, subject to section 363(k) of this title, of any property that is subject to the liens securing such claims, free and clear of such liens, with such liens to attach to the proceeds of such sale, and the treatment of such liens on proceeds under clause (i) or (iii) of this subparagraph; or
(iii) for the realization by such holders of the indubitable equivalent of such claims.49
In contrast to the text of Section 363(k), the text of Section 1129(b)(2)(A)(iii) does not specifically indicate that secured creditors have the right to credit bid. Thus, on several occasions, debtors have filed a reorganization plan providing a cash-only auction-which denies secured creditors the right to credit bid-and have argued that the plan offers these secured creditors with "the indubitable equivalent" of their claims nonetheless.5
Therefore, the question is whether a bankrupt debtor can "cram down" a reorganization plan that denies the secured creditor of the possibility of credit bidding.51 Three separate recently decided cases demonstrate this controversy. Although the Courts of Appeals of the Fifth and Third Circuits ruled in favor of the debtors52 and confirmed a reorganization plan that denied the secured creditors the option to credit bid, the Seventh Circuit sided with the secured creditors and refused to confirm these plans.53 The two circuit courts that ruled in favor of the debtor relied on the Code's plain language (which uses the word "or" to separate Subsection (ii) from (iii)),54 but the Seventh Circuit rejected the argument that Section 1129(b)(2)(A) solves the problem in plain language.55 The court also interpreted the term "indubitable equivalent" to be the face value of the over-secured creditors' claims and the current value of the encumbered asset for the under-secured creditors.56 The court further explained that for the under-secured creditors, the value of the encumbered asset can be either judicially evaluated or auctioned in the free market.57 However, for the latter mechanism to be effective, credit bidding must be allowed.58 The court explained that by allowing secured parties to credit bid, the Code provides lenders with the means to protect themselves from the risk that the winning auction bid will not capture the asset's actual value.59 The court emphasized that this risk is "substantial" because of several features of bankruptcy auctions.60
To fully understand the actual conflict underlying the "credit bidding" legal controversy, one should be better acquainted with the economic surroundings of bankruptcy auctions.
B. Economic Environment
1. Why an Auction?
Not all financially distressed firms require a formal bankruptcy proceeding. Firms often liquidate without undergoing a Chapter 7 proceeding. Many financially distressed firms reorganize or restructure their capital without initiating a Chapter 11 proceeding.61 However, for those financially distressed firms that opt for the sanctuary provided by the bankruptcy court-which manifests in a stay imposed on the creditors' debt collection efforts against the firm -perhaps the most important question concerns redeployment. In both Chapter 7 and Chapter 11, the question is as follows: what should be done with the assets of the financially distressed firm? What is the best way to maximize the value of these assets for the benefit of the firm's claimants (i.e., creditors and shareholders alike)63 as well as the benefit of the economy at large?
As usual, having a crystal clear goal-maximizing the proceeds of the firm's assets-does not prevent vigorous discussions over the "how" (i.e., over the best way to accomplish this goal). For many years, bankruptcy scholars have debated the proper route to which the bankruptcy court should push the financially distressed firm once it has entered a formal bankruptcy proceeding.64 These discussions have been held regardless of the current legal regime dictated by the Bankruptcy Code because the Code is flexible enough to entertain just about any route chosen.65 One route, which may be termed "administrative" on account of its non-market nature, focuses on arranging a pretended sale of the firm to its claimants, who negotiate an arrangement among themselves that must require each of the claimants to provide certain concessions. Although many variations exist, the basic deal is debt swapped for equity. For example, creditors may be asked to forgo part, perhaps even all, of their claims against the firm in exchange for equity in the restructured firm. A relatively lengthy and complicated negotiation process will occur under the auspices of the bankruptcy court because the parties must agree on a value for the firm's assets to ascertain how much the firm's equity is worth and what percentage of this equity the creditors should receive in exchange for giving up their claims. At the end of this process, the parties hope to achieve an arrangement. For example, if the legal framework is a Chapter 11 proceeding, this arrangement must be accepted by at least a supermajority of the firm's claimants, as required by law,66 and confirmed by the court.67
A second route to making a redeployment decision focuses on either "creating" or "locating" the firm's actual "residual owner." These words are in quotation marks because the observation that they reflect relies on an economic analysis. Indeed, from an economic perspective, the source of the redeployment problem lies in the fact that no one actually knows the real value of the firm's assets, if this value can be ascertained at all. Indeed, if one could determine that the assets of a firm that owes $7 million to secured creditors and $7 million to unsecured creditors are only worth $4 million, then obviously the bankruptcy court must announce that the secured creditors are the new shareholders of the firm and must immediately terminate the bankruptcy proceeding. As a result, the new shareholders will need to decide on their own what to do with the firm's assets, which would now carry no debt at all because the secured creditors swapped their claims for equity and because the unsecured creditors' claims were wiped out on account of being "underwater." Obviously, these new shareholders are better motivated than the court to make the best decision regarding the redeployment of the firm's assets and no less knowledgeable than the court in conducting business.69
However, in an imperfect world in which agents are occasionally wealth-constrained and sometimes refrain from incurring even beneficial risks, bankruptcy theory70 and practice converged to recommend that lawmakers seeking a good decisionmaking procedure control the redeployment of financially distressed firms' assets through a simple yet potentially efficient practice: the auction.71 In a typical scenario, the assets of the distressed firms are offered for sale in the free market to whomever is willing to pay the highest amount. In this way, two important decisions are separated and thus prevented from interfering with one another: what to do with the assets to maximize their value and how to divide the value (and how past claims against the firm should be settled). Most importantly, the assets can be auctioned either piecemeal or as a going concern,73 albeit in a relatively quick and cost-economizing manner.74 These features qualify the auction as a satisfactory solution, even for those who believe that the financially distressed firm should be kept alive.75 From the buyers' side, in a sufficiently developed economic environment such as the United States, even cash-constrained buyers can participate in these auctions, as these buyers can employ a variety of financing techniques, including leveraged buyouts (LBO).76
In practice,77 the auction is usually held in the offices of the bankrupt firm's attorneys. When large firms' assets are auctioned, investment bankers may be hired in advance in exchange for a percent of the future proceeds (usually a 1% "success fee") to advise the bankrupt firm in the auction process.78 These advisors disseminate information about the auctioned assets by, for example, preparing brochures, arranging "data rooms," and soliciting bids from prospective buyers.79 Sometimes, these bankers will provide financing options for prospective bidders. An example of a financing option is "stapled finance" (i.e., a loan commitment at prespecified terms to whoever wins the bidding contest, which the winner can accept or reject).80 Following the auction very closely, the bankers may also locate a buyer who commits to purchasing the auctioned assets at a specified price unless the buyer is outbid at the auction. Often referred to as "a stalking horse," this initial buyer is an important figure, mainly because of the significant costs associated with bidding in the auction of a large firm. Indeed, a prospective bidder preparing for the auction must appraise the auctioned assets in advance. Executing this evaluation process (often referred to as "due diligence") is costly; for large firms, the cost may run up to several million dollars.81 Therefore, the stalking horse, whose offer serves as an initial bid and is "shopped around" to elicit other bids,82 is given a guaranteed "break-up" fee (i.e., a reimbursement for its expenditures83) by the bankrupt firm if someone else outbids him in the auction.84 Other protection measures are also sometimes agreed upon between the firm and the stalking horse. For example, there may be an overbid requirement, which sets a minimum price sufficiently higher than the stalking horse's bid for any other bidder who wishes to outbid the stalking horse.85 Of course, these protection measures may influence both second bidders and the result of the auction.86 Thus, the bankruptcy court is called upon to examine these measures, and courts sometimes intervene to change the measures upon which the debtor and the "stalking horse" have agreed.88
Having won the auction and upon obtaining the bankruptcy court's approval for a purchase agreement with the bankrupt firm- as required by law89-the buyer receives the assets free from all past liabilities.90 These liabilities are left with the financially distressed firm, which is now revealed as having traded its assets for cash (i.e., a different type of asset). Thus, a bankruptcy court that allows or mandates an auction of the bankrupt firm's assets is subsequently left with a rather simple task: distribute the proceeds from the sale to the firm's claimants according to the pre-bankruptcy ranking of their priorities. Thus, the auction process generates a new "residual owner" for the assets: the buyer. In both the past and the future, efficiency appears to reign everywhere.
Consider the past first. Ex-post efficiency is obtained if the following three conditions are fulfilled: the value of the bankrupt firm's assets is maximized for the benefit of its claimants, the direct and indirect costs91 spent on executing the bankruptcy proceedings are low, and the pre-bankruptcy entitlements are honored. Having been executed against the background of well-functioning capital markets, the auction can be assumed to have produced the highest revenue possible for the distressed assets92 while consuming few transaction costs on the way. The auction also generally maintains the Absolute Priority Rule because the proceeds from the auction can be distributed among the claimants according to their pre-bankruptcy rankings. In other words, senior creditors receive full payments before any of the junior creditors can see any payments on account of their claims.94 Ex-ante efficiency-an aspect of bankruptcy that focuses on disciplining the behavior of the firm's insiders-is also gained, as the creation of a new residual owner punishes the incompetent shareholders and managers who brought the firm to a state of financial distress.95
As for the future, as a result of the auction, the firm can emerge from bankruptcy with a new capital structure.96 Examining the future also necessitates a close look at the best interests of society in general. In this context, we find that the assets previously owned by the distressed firm are now owned by the one market actor who was willing to pay the highest amount of money for them (this actor is also presumably their highest-valued user). Thus, because the assets have moved consensually to their highest-valued user, the move is efficient. Except for the contexts in which the bankruptcy procedure is expected to pursue redistributive goals,97 the auction seems like the perfect solution.
2. Challenges for Bankruptcy Auctions
Unfortunately, bankruptcy practitioners and scholars quickly discovered that this naive account of the bankruptcy auction fails when faced with reality.
First, capital markets, which are supposed to produce bidders in sufficient numbers, and with a sufficient degree of liquidity to ignite competitive auctions, cannot be currently described as functioning perfectly. Experience and evidence show that too often only a few bidders participate in bankruptcy auctions,98 and that those bidders usually rely on their own existing resources to finance the purchase of the auctioned assets rather than assembling the necessary funds in the markets.99 On many occasions, the bidders in the bankruptcy auction come from the same industry to which the financially distressed firm belongs.100 This fact is significant because it may mean that if the industry itself is undergoing a period of downturn, potential bidders at the auction will be cashconstrained.101
Second, a bankruptcy auction is usually held on a tight time schedule. The claimants of the financially distressed firm want to collect the money owed to them as soon as possible. Creditor pressure increases immensely in the presence of over-secured creditors, who have nothing to gain from postponing the auction in an attempt to conduct it under better conditions.102 However, these creditors do have the leverage needed to pull over the financially distressed firm (sometimes because this creditor also finances operations during the firm's Chapter 11-the "DIP financing") and thus push aggressively for what has become known as a "fire sale."104 Fire sales may be a source of inefficiency insomuch as the speed of the sale undermines the process of shopping around the assets or forces a sale when potential buyers are suffering from temporary illiquidity.105
However, the bankruptcy practitioners involved in the case (as long as they do not represent the unsecured creditors) may benefit from such sales, especially if these sales can be accomplished quickly without having to disclose information to the creditors and confront possible objections from the creditors (if the creditors are consulted regarding the best redeployment path).106 Some argue that these practitioners often have conflicting interests because the buyer may provide them with future business.107 In turn, the practitioners can easily manipulate the auction in that buyer's favor.108 Occasionally, some of the firm's insiders share enthusiasm for a quick sale because their pay is tied to either the accomplishment of such a sale or the interests of particular potential buyers who promise to hire them.109 In this context, it has been argued that competition among bankruptcy courts over bankruptcy cases drives judges to be less adamant in blocking such sales, despite their poor results.110 In some cases, particularly in the bankruptcies of small-to-mediumsized firms, the short schedule is the result of a buyer appearing before the bankruptcy forum and offering to buy the assets, but conditioning the deal on it being accomplished within a short period of time."1
Worse still, unlike a non-bankruptcy sale, the auction cannot, in principle, be postponed to survive periods of macroeconomic downturns, such as a recession in the economy, a general credit shortage, or an industry downturn. The exact purpose of the bankruptcy proceeding is to collapse the future values of the assets to a sum of cash in the present. As a result, those economic declines that do not necessarily concern the distressed firm nevertheless undercut the amounts bid at the auction and the intensity of competition among the bidders.
Third, previous research shows that bankruptcy auctions often attract scavengers. Everybody knows that the auction is run under a deadline and cannot be postponed. If the assets are sold in bulk in an attempt to retrieve their going concern value, bidders understand that a reasonable alternative is a piecemeal sale of the assets, which would generate a significantly smaller amount because no going concern value would then be sold. The fact that the sale is associated with a failed firm may also contribute to the feeling that one can find a bargain in a bankruptcy auction. Occasionally, potential buyers are not sufficiently informed with respect to the real 1 1 ^ value of the auctioned assets. Thus, although the assets are to be purchased in the safest legal way possible-under the auspices of the bankruptcy court, whose involvement actually wipes out any past entitlements that may threaten those of the future buyer-the bids submitted in a bankruptcy auction are systematically skewed toward a lower price. As a result, after the formal completion of the auction, negotiations with bidders are sometimes encouraged by a persistent bankruptcy court, unwilling to accept the poor results of the auction. Such negotiations may result in increased bids.114 The only notable exception concerns the bankruptcies of small-to-medium-sized firms, where the major creditor of the firm is a bank that may also finance the winning bidder at the auction."5 In these cases, the role played by the bank pressures the bids to be higher.116
Fourth, unlike a non-bankruptcy sale of assets, the interests of the formal owner-the firm, particularly its incumbent shareholders and managers-are not always fixed on the single goal of maximizing the proceeds of the sale. Bankruptcy auctions may be managed by a trustee if the bankrupt firm is undergoing a Chapter 7 proceeding. Alternatively, the auctions may be managed by the firm's incumbent management, who serve as a Debtor-in-Possession, if the firm is in Chapter 11. Particularly in the latter case, the manner in which the auction is conducted may be influenced by the fact that the shareholders and managers of the bankrupt firm may have interests other than the simple maximization of value in mind. They may wish for the assets to be sold to a "white knight" (i.e., a buyer who would later cooperate with them).117 Worse still, the shareholders and managers may even appear at the auction house as bidders themselves, albeit disguised and hidden ones.118 The purpose of such a maneuver, which is called a "sale-back" or a "freeze-out," is to become the owners of the auctioned assets once again. However, this time, there will be no liabilities attached to the creditors.119 Obviously, the interest of these disguised bidders is to buy the assets at the lowest price possible.
The fact that the firm's incumbent shareholders and managers may have an interest other than maximizing the proceeds received for the assets in an auction may have several important implications. Whereas some believe that participation by insiders in the bankruptcy auction may convey information to outsiders contemplating the value of the assets and may encourage them to participate and increase their bids in the auction,120 others have argued that insider participation is harmful. For instance, insider participation may exacerbate a preexisting problem of information gaps.1 1 Indeed, not all of the information about the auctioned assets available will be disclosed to the potential bidders. Insiders often have unique information (which may or may not be verifiable) about the firm, its assets and its modus operandi. Consider, for example, a trade secret, the potential profits from an ongoing R&D process, or even a hidden antitrust violation, which, although illegal, may nevertheless remain undetected by the authorities and thus may contribute to the value of the assets. Some of this information cannot be disclosed at all because it may compromise the firm's future operations. Some of this information is non-verifiable, and there is no value attached to disclosing it anyway. Finally, some of this information may be hidden from potential bidders by interested insiders.
Moreover, the possibility of disguised insiders participating in the auction can generate a "chilling effect" on the competition. Aware of the possibility that one of the bidders is actually an insider in disguise, other bidders fear the possibility of winning the auction.122 Known as "the winner's curse,"123 these other bidders fear the prospect of submitting an excessively high bid. After all, they may win the auction while outbidding an insider, who is assumed to be better acquainted with the auctioned asset, because they misevaluated the asset. This logic causes bidders to either refrain from participating in the auction altogether or submit lower bids to increase the chances of winning the auction with a bid that is low enough to guarantee a good bargain.
A slightly different chilling effect in the context of insider participation concerns over-bidding. The incumbent management may join a creditor of the firm to form a coalition that will bid in the auction.124 In this case, this coalition has an incentive to overbid (i.e., bid above its true valuation of assets) to push for a higher counteroffer from the other bidders because these counteroffers will eventually become the payment for the coalition's pre-bankruptcy stake in the assets.125 In short, these insiders are caught in a conflict between their interests as claimants and their interests as bidders. As a result, the coalition may win the auction, and the assets will not be assigned to their highest-valued user.126 Furthermore, outside bidders may be deterred from competing in such auctions in the first place, making these auctions less competitive.127
At the end of the day, the relative efficacy of bankruptcy auctions remains controversial, particularly in comparison with the alternative: the reorganization and restructuring of the firm. The gathered empirical data have also been unable to resolve the debate between those advocating the bankruptcy auction as the tool of choice128 and those arguing that auctions in bankruptcy should be mistrusted or at least handled with caution.129
II. Auction Theory-A Primer
Although an exhaustive discussion of the auction theory literature is far beyond the scope of this Article,130 the purpose of this Part is to provide the reader with a short introduction to the areas of auction theory that are relevant to our issue.131
A review of the auction theory literature shows that for any given set of circumstances, a different type of auction will produce a different result. Thus, any decision on the type of auction to be used in a particular set of circumstances and the set of rules to be applied will impact the revenue that will be generated by that auction.
Auction theory scholars seek to answer the following question: given a particular factual situation, which allocation mechanism will maximize the revenue for the seller? In other words, what is the optimal auction design?132 The environment in which the auction is held is characterized as one of uncertainty in the sense that the parties do not know the market value of the asset prior to the execution of the sale. In addition, even if they have determined its value from their own perspectives, they do not know its value to the other players.133
Auction theory also considers the other rules (in addition to those that are a given aspect of the chosen auction type) that should be considered by the auction designer. For example, is it efficient to set a reserve price for the auction? What is the proper disclosure policy with respect to the relevant information known by the seller? These and other questions regarding optimal auction design will be discussed below.
Since the 1980s, hundreds of theoretical and empirical studies have investigated the subject of auction theory.134 The insights developed in these studies have also been implemented on a practical level in many governmental auctions, some of which were designed with the help of auction theory scholars. Some of these auctions were very successful, whereas others have failed. Both the successes135 and failures136 have been attributed to the design of the relevant auction.
?. A Typology of Auctions
We begin by presenting the main types of auctions analyzed by auction theory that can also be relevant in the context of a bankruptcy auction.137 Theoretically, there is a wide range of possible auction formats. However, the discussion in the auction theory literature focuses on four basic types of auctions.
The first type of auction is the sealed-bid auction, which is also called a first-price auction.139 This auction primarily occurs in the context of government procurement. In this auction, each bidder submits a bid in a sealed envelope. All of the envelopes are opened at the same time. The bidder who has submitted the highest bid wins the auction and pays the amount indicated in her bid.
The second type is an English auction or an ascending-price auction.140 In this format, the auctioneer begins by announcing a low amount for the price. The amount increases as the bidders continue to raise it. Eventually, only one bidder remains, and this bidder is the winner of the auction. The price paid by the winner is the amount of the last bid that she submitted. Of course, English auctions can also be conducted online, and there are also various sub-types of English auctions; for example, the identities of the bidders can be known or kept secret.141
The third type of auction is the Dutch auction, which begins with the auctioneer announcing a price higher than the market value of the asset.142 Immediately afterwards, the price begins to drop, gradually and continually, until one of the bidders signals to the auctioneer to stop. At this point, the auction ends; the bidder who has stopped the auctioneer is the winner and pays the price that was called by the auctioneer at the stop. This type of auction is called a Dutch auction because it is the form used in the Dutch flower market. Dutch auctions can also be conducted through an electronic presentation of the continually decreasing price. In this format, each bidder can press a button to stop the auction.
The fourth type of auction is the second-price auction, which is conducted in the same manner as a sealed-bid auction in that the bidders submit their bids in an envelope.143 The winner is the highest bidder. However, unlike the sealed-bid auction, the amount paid is not the amount offered by the highest bidder. Rather, the amount offered by the second highest bidder is paid. This auction is also sometimes called a Vickrey Auction, which was named after the economist who first described the format in his seminal article published in 1961.144
The literature refers to other types of auctions beyond the four basic types described above.145 For example, an interesting auction format is the "Anglo-Dutch auction," in which an English auction is conducted until the number of bidders remaining in the competition is one more than the number of items that are being auctioned. At this stage, the English auction ends, and the bidders are invited to submit a single-sealed bid, which may not be lower than the last bid made in the English auction. The Anglo-Dutch auction is supposed to combine the best of both worlds by mixing elements of the English auction and the sealed-bid auction. This type of auction offers the following advantages of an English auction: a simple strategy, the ability to gather information during the course of the auction, and reduced risk of the winner's curse. Moreover, it encourages weak bidders, who hope to do well in the second round, to participate in the auction and galvanizes bidders to violate any existing cartel agreements.146
C. Choosing an Auction Method
Before we continue our search for the optimal auction design, we should discuss the characteristics that are common to the auction types mentioned above and the connections that exist between them.
1. Sealed-Bid Auction and Dutch Auction
From the perspective of auction theory, the sealed-bid auction and the Dutch auction are completely identical. The set of considerations and the strategy to be weighed by the bidder is the same in these two auction types.147 We will first examine the bidder's considerations in the sealed-bid auction. Given that the bidder values the auctioned asset at X and that the bid that she submits is Y, the profit the bidder expects to receive from the auction is equivalent to X less Y. When the bidder formulates her bid, she faces a dilemma. On the one hand, the higher the bid, the greater the chance that she will win. On the other hand, the higher the bid, the lower her profit from the auction will be, and vice versa. Therefore, the bidder must weigh these conflicting interests in submitting her bid.
The same set of considerations comes into play in a Dutch auction as well. As the auction progresses, the bidder debates when to stop the auctioneer. The longer the bidder waits, the greater her expected profit will be, although the probability that another bidder ends the auction before she does will also increase accordingly. The set of considerations that the bidder faces in this situation are the same as those faced by a bidder in a sealed-bid auction. Thus, in a given situation, the bidders will behave in an identical fashion in the two types of auctions, and the two formats should therefore yield the same results. For the sake of convenience, we will refer from this point forward only to the sealed-bid auction, but unless otherwise indicated, the points raised in the discussion will be true for a Dutch auction as well.
2. English Auction and Second-Price Auction
The English auction and the second-price auction also share characteristics, and except for certain qualifications that are discussed below, auction theory views the two formats as producing the same result. In addition, the English auction mechanism and the second-price auction have an important common characteristic: under any set of circumstances and in any environment, they both ensure that the auctioned asset will be allocated to the party that values it the most. The same cannot be said with respect to the sealed-bid and Dutch auctions.
First, we will show that the English auction and the secondprice auction both ensure that the asset is allocated to the bidder who values it most highly. Assume that bidder A values the asset at 100, bidder ? values it at 95, and bidder C values it at 90. A purchase of the asset at a price higher than the bidder's valuation will not be worthwhile for that bidder, as such a purchase will constitute a loss transaction for her. Thus, the English auction will be conducted as follows: as long as the auctioneer calls a price that is lower than 90, all three bidders will remain in the game. If the price reaches 90, bidder C will leave the competition, as a purchase at 90 is no longer worthwhile to her. Any bid above 95 is not worthwhile to bidder B. Thus, once the auctioneer reaches 95, bidder ? will leave the auction and bidder A will purchase the asset at 95. Ultimately, the asset has been allocated to the party who values it the most (bidder A). This bidder will pay the price at which the next highest bidder (bidder B) values it. This result is stable and is not sensitive to a change in circumstances. In game theory terms, the English auction provides the bidder with a dominant strategy equilibrium; each bidder has a well-defined optimal strategy that does not depend on, and is not connected to, the bids submitted by the other bidders. Specifically, this strategy demands that the bidder remain active in the auction as long as the price is lower than the price at which the bidder values the asset. Conversely, the bidder leaves the auction if the price reaches the amount at which she values it.
In a second-price auction, the amount that will be bid by a rational bidder is also the price at which she values the asset. Thus, the asset will be allocated to the bidder who values it the most, and that party will pay the price at which the second highest bidder values the asset. 48 For example, assume that there are four bidders competing in a second-price auction. Bidder A values the asset at 90, bidder ? values it at 94, bidder C values it at 95 and bidder D values it at 100. None of them know the other bidders' valuations. Let us look at the strategy of bidder B. Bidder ? has no interest in bidding less than 94 because placing a lower bid would not provide her any benefit. It reduces the bidder's chances of winning and does not increase her profits even if she wins because the bidder will not determine the amount to be paid. That amount will be determined by the bidder making the next highest bid (bidder A). Bidder ? also has no interest in bidding more than 94 because even though she will increase her probability of winning by doing so, the price at which she would achieve her victory would not be in the range within which she wishes to win. By increasing her bid beyond 94 (e.g., to 96), she will be able to "beat out" bidder C, who will bid 95. Thus, she may win an auction that she would otherwise have lost. However, in this situation, it would be preferable for her to lose because her winning will necessitate that she pay 95 for an asset that she values at only 94. Therefore, she will always bid 94, the price at which she actually values the asset. This strategy is the dominant strategy.149 Thus, in the example given above, bidder D, who values the asset at 100, will win and will pay 95. This result is completely identical to the result of an English auction.150
As noted above, the English auction and the second-price auction both offer the bidder a dominant strategy. Thus, the individual bidder's strategy will not change regardless of whether the bidders are risk-averse or risk-neutral. Moreover, it will not matter that only one asset or several assets are being auctioned. Similarly, the issue of whether there are few or many bidders will not change each bidder's strategy. These and other variations, which could have a significant effect on the bidders' strategies in the case of a sealedbid auction, will not change the bidders' strategies in an English auction or in a second-price auction. Therefore, they will not have any impact on the final allocation of the asset.
3. Why Are Second-Price Auctions So Rare?
In light of the above, the following question arises: if an English auction and a second-price auction are effectively the same "game," why are second-price auctions so rare?151 An in-depth discussion of this issue is beyond the scope of this Article. However, we will briefly note two main reasons for this phenomenon.
First, there is an increased incentive for the seller in a second152 price auction to perpetrate fraud to increase her revenue. The danger is that after the envelopes are opened, the seller will submit a fictitious bid that is slightly less than the highest bid. This bid will not affect the identity of the winner but will increase the amount that she will pay. As long as this danger exists, the bidders' best strategy is to lower their bids below their actual valuations of the asset and thereby deviate from the theoretical dominant strategy. This strategy will be optimal even if the fraud does not actually occur because the mere presence of the danger of fraud will be sufficient to create a perceived need on the bidders' part to change their strategies.
Second, auctions categorized as second-price auctions are rarely used because they expose the size of the winning bidder's expected profit.153 Specifically, when the auction ends, all of the participants discover the amount that the winner would have been willing to pay for the asset (i.e., the amount of her bid), what the winner will actually pay for it (i.e., the amount of the second highest bid), and the winner's expected profit (i.e., the difference).154 The exposure of this information is problematic from at least two perspectives. First, if the winner wishes to perform work involving the asset or to sell it to a third party, any party with which the winner negotiates will know the limits of her flexibility. The improved bargaining position of the other side will harm both the winner and the seller because the bidders, who will be aware of the forthcoming scenario in the event of their victory, will consequently submit lower bids.155 Second, an additional weakness of the second-price auction is that a large gap between the winner's valuation of the asset and the price that is actually paid could put the seller in an embarrassing position and expose her to criticism.156
In sum, a second-price auction would appear to be an undesirable mechanism at first glance. However, at second glance, the second-price auction has several advantages, which raises the question of why the mechanism is used so rarely. A further review provides the answer by discerning various weaknesses involved in using this type of auction as an allocation mechanism. Consequently, in the following discussion, we will focus on the English auction, which will be compared with the possibility of using a sealed-bid auction.
4. Sealed-Bid Auction Versus English Auction
After presenting the main characteristics of, and the connections between, the four basic models for auction mechanisms, we turn to the following question: which format is preferable from the seller's perspective? As will be shown below, there is no simple answer to this question. The correct answer is that, in certain circumstances, an English auction will be more efficient, but in other circumstances, the sealed-bid auction is preferred. We will now present some of the characteristics, features, and circumstances to be considered when choosing the preferred auction mechanism in a particular situation. We will first present the features or circumstances that give an advantage to the English auction mechanism. Afterwards, we will present the same analysis for the sealed-bid auction.
a. When Is the English Auction Preferable?
1. Transaction Costs and Erroneous Strategy
The optimal bidding strategy in an English auction is relatively simple. The bidder knows her valuation of the asset and knows that the transaction will be worthwhile as long as the auction price does not exceed that value. There is no incentive for the bidder to engage in industrial espionage against the other bidders, as she will gain no advantage from doing so. The information that the bidder needs to formulate her strategy-the identity of the other competitors and their valuations of the asset-will be exposed as the auction proceeds. This information leads to smaller transaction costs because there is no need to engage in industrial espionage or to defend against such espionage from other bidders. Accordingly, a bidding strategy can be formulated in a relatively simple manner.
The bidder's situation in a sealed-bid auction is different. In a sealed-bid auction, the formulation of each bidder's strategy depends on what information she possesses regarding the other bidders' identities and strategies. This dependency creates three clear disadvantages. First, it obligates the bidder to gather information about her competitors, which increases the transaction costs. Second, the bidder will also want to take action to protect herself against industrial espionage from the other bidders. This protection measure also increases the transaction costs. Third, industrial espionage involves the risk of a mistake or a failure in the gathering of information. As a consequence of these factors, overall efficiency declines. This feature gives a significant advantage to the English-auction mechanism over the sealed-bid auction.
2. Incomplete Information and the "Winner's Curse"
As a starting point in auction design, the seller and the bidders will usually have incomplete information regarding the value of the asset being sold. One of the phenomena that characterize this situation is the winner's curse.157 In this event, the auction winner eventually discovers that she valued the asset at a higher value than its market value such that her victory in the auction has caused her to enter a loss transaction. The explanation for the winner's curse phenomenon is that in circumstances of incomplete information, there is a reasonable chance that the bidders will make errors-both upwards and downwards-regarding the asset's true value. However, although a downwards error is not excessively costly (i.e., the bidder simply does not win the auction), an upwards mistake can be very expensive for the bidder, as she will be forced to enter a loss transaction. Naturally, the auction's winner will always be the most 1 SX optimistic bidder, but her optimism may have been excessive.
How can the bidders deal with the winner's curse? The answer is that in situations characterized by uncertainty regarding the assets' value, each bidder must assume that her bid will be the highest and that she may be overly optimistic and may have overvalued the asset. To reduce the risk of a win that results in a loss transaction, the bidder must reduce her valuation by some degree. Because the size of this reduction can be difficult to calculate,159 uncertainty regarding the asset's value works against the seller as well and provides her with an incentive to reduce the bidders' uncertainty about the asset's value as much as possible.
Milgrom and Weber have shown that from the perspective of the winner's curse, the English auction is to be preferred over the sealed-bid auction160 because the English auction mechanism provides the bidders with information regarding the other bidders' valuations of the asset.161 Consequently, the bidders in English auctions are less concerned about falling victim to the winner's curse and are willing to bid higher amounts.
As an interim conclusion, we note that, with a few exceptions, the expected revenue for a seller conducting an English auction will be higher than the expected revenue from a sealed-bid auction. We will now discuss the circumstances in which a different conclusion may emerge.
b. When Is the Sealed-Bid Auction Preferable?
In an English auction, the risk-averse nature of the bidders will be of no consequence. However, this is not the case in a sealedbid auction, in which the fear of losing will give a risk-averse bidder an incentive to submit a higher bid that is closer to her actual valuation. By doing so, the bidder increases her chances of winning even if the bidder's expected profit is consequently lower. Thus, in a sealed-bid auction, the more risk-averse the bidder is, the closer her bid will be to her valuation of the asset and, accordingly, the higher that bid will be.163
The primary enemy of an auction is a bidding ring or cartel. The objective of a cartel is to frustrate the auction process by creating a false impression of competition, when in actuality, the cartel members have already agreed amongst themselves on how each of them will behave (i.e., which member will win the competition, how high that member's bid will be, and how the excess profit will be distributed among the cartel members).164
However, a cartel is not a stable organization. Cartels have a tendency to collapse because each cartel member has an incentive to betray her colleagues and to participate in the auction in violation of the cartel's agreements. Milgrom has shown that the likelihood of cartel activity is greater in an English auction than in a sealed-bid auction.165 As the English auction is conducted openly, the submission of a bid that does not conform to the cartel's mutual agreements will be exposed immediately. This bid will signal to the other cartel members that the cartel's agreements are no longer in force. Under these circumstances, the members will understand that the cartel has dissolved, and they will return to competing with one another. Thus, in an English auction, a potential defector knows that she will not realize any benefit from an act of defection and therefore has no incentive to commit such an act. In other words, the English auction mechanism reduces the incentive for cartel members to deviate from the cartel agreements and thus strengthens the cartel's stability. In contrast, in a sealed-bid auction, no one can know who is participating in the auction and what their bids are until the auction has ended. This situation encourages deviations from the cartel agreements and therefore reduces the likelihood that a cartel agreement will be created or will continue to exist.
Parenthetically, a relatively simple and effective means for reducing the risk of cartel formation is to establish a reserve price that is equal to or slightly higher than the seller's ion of the asset. In this way, the seller protects herself from transaction and also greatly reduces the incentive for collus tivity.166 It would also be advisable for the seller to hide the number and identity of the active bidders to make it more difficult for the bidders to transmit information to each other.167 Consequently, if the auction is conducted under circumstances that create a high risk of cartel activity, the preferred mechanism is a sealed-bid auction.168
Given that participation in an auction involves transaction costs, a bidder will be deterred from participating in an auction if she estimates that her chances of winning are low. In a market characterized by asymmetric bidders,169 the weak bidders are aware that a strong bidder who is determined to win the auction can always submit a bid that will be more attractive than their bids. If the auction is an English one and the bidders are asymmetric, a weak bidder knows that for any bid that she submits, the strong bidder can always add one more dollar until the strong bidder eventually wins the auction. The strong bidder, who is also aware of this fact, has an incentive to signal to the weak bidder that she is determined to win the auction''iand>fhat she does not intend to give up. This type of message can "achieve its purpose and deter the weak bidder from participating in the auction altogether. Such deterrence will minimize the number of bidders, reduce competition, and diminish the seller's expected revenue.170
In contrast, in a sealed-bid auction, the strong bidder does not know how the weak bidder will behave (i.e., what the weak bidder's bid will be or whether the bidder will even participate in the auction). This situation encourages the weak bidder to surprise the stronger bidder and causes the strong bidder to avoid taking any risks by submitting a higher bid. These features can generate an effective level of competition and will consequently be expected to increase the expected revenue for the seller.171 Thus, in an environment characterized by asymmetric bidders, the seller should prefer a sealed-bid auction.
In sum, no sweeping answer can be given to the question of which auction mechanism is better from the seller's perspective. All that can be said is that under "normal" circumstances, English auctions will be preferable. This preference exists because under an English auction, the transaction costs are lower, the bidders have simpler strategies, the winner is less exposed to the winner's curse, and the risk of market failure is also lower. Nevertheless, this conclusion is not valid if an analysis of the market indicates that the potential bidders are significantly risk-averse, asymmetric, or at risk of developing cartel activities. If any of these conditions are present, there are good reasons for preferring a sealed-bid auction.
D. Additional Aspects
Beyond the question of which auction mechanism should be preferred under specific circumstances, additional considerations and possibilities for auction rules must be discussed. These considerations will be relevant regardless of whether the choice has been made to conduct an English auction or a sealed-bid auction. We will now discuss some of the considerations that are most relevant to bankruptcy auctions.
1. Information on the Asset
One interesting question that has engaged auction theory scholars is the determination of the best strategy with respect to the information known by the seller. In many cases, the seller has private information regarding the potential revenue that may be generated by the asset being sold, information regarding restrictions on the use of the land being auctioned, or information regarding the mechanical condition of a machine being auctioned. Milgrom and Weber have shown that, for any type of auction, a full disclosure policy is preferable because this policy will maximize the expected revenue for the seller.172
Although this conclusion is intuitive with respect to positive or neutral information concerning the asset, it seems to be somewhat less logical if the information under discussion is negative. Milgrom and Weber have mathematically proven their argument in favor of full disclosure, but there is also an intuitive explanation based on the rationale for the winner's curse for this conclusion. As noted above, the starting point for understanding the winner's curse is that the bidder does not have complete information regarding the value of the asset. As a result, the bidder must establish her own valuation. The assumption is that a rational bidder who has incomplete information and is aware of the winner's curse will reduce her bid. In other words, there will be a certain amount, which can be defined as a "risk premium," at which the bidder will reduce her planned bid to avoid falling victim to the winner's curse. In such a situation, any additional information that will diminish the level of uncertainty regarding the asset's value will also reduce the level of the risk premium. Thus, the bidder will be willing to submit a higher bid. Although in the case of negative information, the disclosure will not necessarily increase the amount of the bids, it will not reduce them either. Therefore, in a situation in which the bidders have incomplete information, the seller's optimal strategy is to disclose all of the information that she has regarding the asset being sold and to diminish the gaps in the information available to the bidders as much as possible.
2. Reserve Price
Various questions arise regarding the issue of a reserve price. There is a consensus that the establishment of a reserve price is an efficient strategy for the seller,174 if only because this price plays a key role in preventing bidder collusion. From this perspective, the optimal reserve price is the seller's valuation of the asset. It will not be worthwhile for the seller to sell below this price, whereas the transaction will be a positive one if the price is above the seller's valuation.
It can be argued that the establishment of a reserve price will not only prevent loss transactions but will also increase the seller's expected revenue under certain circumstances. For example, it is easy to see that in an English auction, the use of a reserve price that is higher than the second highest valuation but lower than the highest valuation will lead to increased revenue for the seller. The problem is that it is difficult to estimate the difference between the highest and the next highest valuations. Worse still, setting an excessively high reserve price will lead to a situation in which no transaction is performed at all, even though the highest bid would have been worthwhile to the seller.
The theoretical analysis shows that from the seller's perspective, the optimal reserve price is a price that is slightly higher than the seller's valuation of the asset.17" However, this insight is generally not implemented in practice. In reality, reserve prices do not exceed the seller's valuation of the asset, and in many cases, they 176 are even lower than this valuation.
What is the optimal strategy regarding the disclosure of the reserve price? As noted previously, Milgrom and Weber have proven that in most environments, the seller's optimal strategy is to disclose all of the relevant information of which she is aware, including the 177 Â· reserve price. However, in reality, sellers tend not to disclose the reserve price. According to Vincent, a non-disclosure policy regarding the reserve price is understandable and justifiable because disclosure reduces the number of auction participants by eliminating all of the potential bidders whose valuations are lower than the 1 78 reserve price. Consequently, the level of competition declines, and the seller's expected revenue is reduced. Conversely, if the reserve price is unknown, more bidders will participate in the 179 auction.
E. Auction Theory in Action-The FCC Spectrum Licenses Auctions
One of the most significant criticisms directed at game theory is that it cannot be implemented in real-world scenarios. This limitation primarily exists because the theoretical research relies on several assumptions that often do not hold true in the real world. For example, the basic assumption of game theory is that decision makers act rationally. However, numerous studies conducted over the last forty years have shown that this assumption is often of 1 ÎÎ dubious validity. In this respect, auction theory presents a marked improvement over other theoretical analyses. Beginning in the 1990s, auction theory economists have shown that an auction conducted in accordance with the insights developed through theoretical research produces greater revenue, but the auction designer must address practical considerations as well.181 We will demonstrate this point through a brief review of the auctions held since the 1990s regarding the allocation of licenses for the use of frequencies along the electro-magnetic spectrum.182
The FCC is responsible for allocating spectrum licenses for various communications purposes within the United States. After previous attempts to allocate spectrum licenses through an administrative process and a lottery failed,183 Congress decided to allocate the spectrum frequencies through the use of auctions in 1993.184
As a result of the legislation, the FCC was required to decide a number of matters related to the manner in which the auction would be conducted. For this purpose, the FCC engaged the services of the best theoreticians from the field of auction theory. This moment was their time to shine, but the project also put their theories to the test: the auction theory scholars now had the opportunity to prove that the insights derived from the theoretical research could serve as practical tools in the hands of policymakers. Given the complexity of the spectrum license auctions, the many problems that arose in the context of these auctions and their unprecedented scope, * 18S the experts' task was not at all simple.
The auction designers had to make a series of difficult decisions. First, the designers had to decide whether to use an English auction or a sealed-bid auction format. Considering the various characteristics of the different mechanisms, the FCC chose to use a unique type of English auction: a "simultaneous ascending bid auction" with discrete bidding rounds. This auction was composed of many rounds of sealed-bid auctions, with the highest bidder at the conclusion of each round being declared its winner. At the end of each round, the FCC would propose that the bidders undergo an additional round. The auction would end when no bidder was interested in making a bid that was higher than the highest bid of the previously concluded round. In this way, the FCC hoped to accomplish the following: (1) to enable the flow of information among bidders; and (2) to create obstacles that would prevent Î¹ Q/r collusion by preserving the bidders' anonymity. Because this new format had never been tried before, the FCC made a bold move by using it in an auction of such scope and importance.
Second, the FCC needed to decide whether it would conduct a single, simultaneous auction in which all of the licenses would be sold; whether it would be preferable to have a separate auction for each license; or whether the agency would choose some combination of the two methods.187 The main element that the FCC needed to consider in this regard was the synergetic connections among the licenses.
The experts agreed that it would be best to conduct a certain type of simultaneous auction, but they disagreed on the details of the auction188. Additionally, a concern arose that a simultaneous auction could be too complicated for the bidders. In turn, this excessive complexity could cause the auction to fail. In the end, the Milgrom and Wilson proposal was adopted.189 This proposal included six auction rules. First, a continuum of licenses would be allocated in each process, with each license being the subject of a separate auction. Second, the auctions would be conducted simultaneously, with the bidders being able to submit bids for all of the licenses in each round, although the number of bids that each bidder could make would be limited to a specific number of licenses that the bidder had established in advance. The bidders were also required to deposit large bid bonds for each license that they bid upon. Third, each auction would remain open until the auction for the last license ended. In other words, these auctions would all remain open until the round in which no bid was submitted for any license closed and until each bidder could move from auction to auction without limitation in accordance with her efficiency considerations. Fourth, each bidder was required to be active in each round. A bidder who was not active was considered to have withdrawn from the entire auction and could not return to it. This condition was established as a requirement such that the auction would progress at a reasonable pace and the bidders would not be able to "sit on the fence" waiting for the other bidders to submit their bids. Fifth, the FCC retained the right to use its discretion to determine the minimal rate of progress for each auction. Additionally, the FCC had the authority to change that pace during the course of the auction in accordance with the auction's progress. Sixth, to prevent fictitious bids, the FCC stipulated that a bidder would be subjected to a fine if she withdrew her bid and that the fine would consist of the difference between the bid that was being withdrawn and the winning bid. If a bid were withdrawn after an auction had ended, the fine would be 3% greater.190
An additional issue that needed to be decided was whether to establish reserve prices for the licenses. According to the theoretical literature, a reserve price would be advisable if there were concerns regarding weak competition, a limited number of bidders, or cartel activities. In the FCC case, it was anticipated that there would be aggressive competition for the licenses. Therefore, the likelihood of cartel activity was expected to be minimal, and the FCC decided not to establish a reserve price.
Another interesting factor that was considered was the complexity of the auction rules. The rules needed to be comprehensible to the policymakers at the FCC and to the management at the companies that would be participating in the auction such that they would be able to understand the auction rules and the optimal strategy to be derived from them.191 The designers of the auction assumed that if the rules were too complex, the rules would not be understood, which would lead the policymakers to reject them. Therefore, it was agreed that it would be preferable (to a certain degree) to use simple rules rather than complex ones, even at the expense of some of the auction's efficiency.192 As noted above, the final decision was to use a simultaneous ascending bid auction, although this mechanism was completely new and untried.
The first spectrum licenses auction allocated only ten frequencies of three types at a relatively low cost. It is interesting to note that there was almost a complete consensus among all of the economists who were involved in the auction design-both those from the industry and those from the government-with respect to the determination of the auction rules.19
The first auction was held at the end of July 1994 over the course of five days. The auction lasted forty-seven rounds and did not require any intervention from the FCC.1 4 The FCC's boldness and daring appeared to have paid off. The auction was viewed by all as a success, and the revenue that it generated was ten times the preauction estimations.195 Beginning with the 1994 auction, all spectrum licenses have been allocated through auctions,196 although some difficulties have arisen along the way.1
Thus, on the first occasion in which auction theory was given such a prominent role in designing a real-world auction, the results were surprisingly positive. Although this electro-magnetic spectrum license auction was so complex and complicated that no theoretical model could have provided a faithful description, the use of and reliance on auction theory principles and insights were essential components of the auction's success.
F. The Lesson Learned
Auction theory is tested primarily by the degree to which its insights can be implemented in real-world situations. Until the FCC's spectrum license auctions in the 1990s, the truth of the insights described above had only been proven in theory. The FCC auctions broke from the mold by providing indisputable proof of the practicality of the theoretical analysis. Since then, many auctions have provided additional evidence of this real-world practicality.198
It is now undisputed that an auction's design can and does have a determinative impact on its results. An auction's design will affect the number of bidders who will participate in the auction, the quality of those bidders, the amount of the bids, the proper operation of the auction itself, and-most importantly-the revenue generated for the seller. All of these effects have been shown through theoretical analyses, laboratory experiments, and empirical research on various auctions held in the past.
Nevertheless, we cannot draw any unequivocal conclusions regarding the optimal auction because no mathematical formula can fully express the characteristics and circumstances of every possible situation.199 McAfee and McMillan have therefore argued that the primary importance of theoretical auction theory research lies in its development of proper intuitions and the ability to understand the impacts of various auction rules on the behavior of bidders in a particular auction.200 Klemperer argues that the most important consideration in auction design is the encouragement of competition, which can be accomplished primarily by encouraging participation in the auction and by preventing collusive activities among bidders.201 Milgrom agrees with this view and also notes the importance of simplicity in the design of the auction rules.202 All of these scholars agree that the concepts and principles developed by auction theory are of great value in the achievement of all these objectives.
III. Designing an Optimal Bankruptcy Auction
Having understood the legal and economic background of bankruptcy auctions (in Part I) and the basics of auction theory (in Part II), we must now apply the insights offered by auction theory to some of the questions evoked by bankruptcy auctions. Of course, resolving all of these questions is beyond the scope of this Article. The purpose of Part III is to demonstrate the manner in which auction theory can be employed to improve bankruptcy auctions. Section A starts by emphasizing several general guidelines that should be considered when designing a bankruptcy auction. This Section offers a new auction design that should arguably be employed instead of the prevailing conventional design. Section ? analyzes the question of "credit bidding." Section C concludes with a note on the institutional environment of bankruptcy auctions and the need to increase the involvement of the bankruptcy courts in the design of these auctions.
A. General Guidelines
Bankruptcy law does not explicitly prescribe a set of concrete rules for conducting an auction that would maximize revenues and minimize costs. However, both the auction design and the result of the auction must receive court approval. Moreover, as demonstrated in the context of "credit bidding," certain sections of the Bankruptcy Code require, even if only indirectly, a careful examination of the design of the bankruptcy auction.204
1. Utilizing Auction Theory
We argue that when considering any proposed arrangement that includes an auction in a bankruptcy, the bankruptcy court should be aware that the design of the auction can have a determinative impact on the likelihood that the auction will achieve its goals and that the judicial authority must ensure that the particular rules of the auction incorporate the insights provided by auction theory. To the extent that practical conclusions can be drawn from economic research, these conclusions should also impact the legal mandate formulated by the court to those in charge of executing the auction. For example, one of the main criteria for obtaining court approval of an arrangement-whether for the purpose of liquidation or reorganization-should be the maximization of the revenue from the auction. Thus, if it becomes apparent that a specific auction design can be expected to generate more revenue than a different design in a particular situation, the approval of the arrangement should be conditioned on the use of the most efficient design under the circumstances.
However, what is the optimal bankruptcy auction design? First, as in many other design contexts, one size does not fit all. Thus, although a default auction design may be suggested, lawmakers should understand that in certain cases, the default auction design should be altered to fit the special needs of a particular bankrupt firm. For example, an auction that cannot be characterized as a "common-value auction" should be accorded special consideration.205
However, having reviewed some of the insights offered by auction theory, we are confident that a default auction design can be formulated. The following is an attempt to offer such a design.
2. The Suggested ADVA Design
Although separate sales of standard assets should also be of interest to lawmakers contemplating an optimal auction design, the more interesting case concerns those bankruptcies in which the auctioned "asset" is the entire business as an ongoing concern. Usually, the large amount of money at stake in these cases, alongside the complexity and opacity of this "asset," merits a careful planning of the auction. However, it is assumed that the auctioned asset is tradable and of common value to all of the potential bidders in the auction.
Broadly speaking, a bankruptcy auction should be designed to maximize the proceeds (revenues, in auction theory jargon) from the sale while minimizing the transaction costs and protecting the interests of the various claimants. With regard to maximizing the revenues and minimizing transaction costs, auction theory fits the context of bankruptcy auctions perfectly because the theories and research executed by the scholars working in that field target these same goals. However, the bankruptcy auction should also protect the interests of the bankrupt firm's relevant claimants. For example, the line of cases concerning "credit bidding" emphasizes the need to protect the interests of a secured creditor holding an under-secured claim if this creditor faces an oppositional debtor-in-possession. Conversely, the interests of junior claimants, including unsecured creditors and equity holders, should also be protected if, for instance, a secured creditor holding an over-secured claim cooperates with the debtor-in-possession or with the firm's shareholders in a manner that may "squeeze out" the interests of the unsecured creditors. In fact, in the context of bankruptcy auctions, one should bear in mind that a conflict of interests is always expected once the claimants are ranked by priority. The uncertainty with regard to the true value of the assets to be auctioned exacerbates this conflict of interests.
The economic environment in which bankruptcy auctions usually occur includes several unique features that are sometimes interrelated and that should be taken under consideration. For the most part, bankruptcy auctions are usually conducted in an environment that can be characterized as relatively uncompetitive, as
few bidders usually participate. The causes of this relatively uncompetitive environment may include one or more of the following: first, the auction is oftentimes conducted during an economic downturn and cannot be postponed; second, the sale of the auctioned asset should be concluded swiftly, as the claimants, especially the creditors, want to be paid quickly; and third, for the potential bidders, the value of the auctioned asset is quite uncertain because of the financial distress context.
Thus, lawmakers should opt for an auction design that would encourage competition and prevent possible cartels among bidders. This suggests an optimal default design that is quite different from the one currently employed in many bankruptcy auctions. The default model design that we envision for large firms sold as an ongoing concern-bearing in mind that some special cases may merit a different approach-is a design that we call the "AngloDutch Veto Auction" (or in short: "ADVA"). An ADVA design consists of an Anglo-Dutch auction without a reserve price and without a pre-contract with a "stalking horse." At the end of the auction, if the bidding price does not exceed the amount of the secured claim, the secured creditor should be extended with a right to veto the sale in exchange for paying a pre-defined amount as a cost to the highest bidder. Consider the following explanation of this design.
a. The Anglo-Dutch Feature
Generally speaking, in the context of bankruptcy sales, a process that revolves around an auction-like interaction with potential buyers is superior to a process that involves negotiations with them. Such negotiations consume time and undermine competition among bidders because potential bidders may fear that these negotiations, which are conducted discretely and may remain undisclosed, have been manipulated to favor those bidders with personal or other connections to the bankrupt firm or to the debtorin-possession. Thus, both a negotiation with a "stalking horse" and a negotiation with the highest bidders (after the conclusion of an initial bidding process) are unwarranted.207
An open public auction is superior to a sealed-bid auction because an open public auction eliminates the need to spread information among the bidders and decreases the possible effects of a "Winner's Curse." An open public auction helps achieve these goals because, as the auction evolves and the submitted bids increase in amount, each bidder draws positive signals from knowing that she is not the only one bidding high. In contrast, a sealed-bid auction is incapable of increasing certainty in this manner among the potential bidders, and the bidders may prefer to bid low to avoid the "Winner's Curse."
An Anglo-Dutch design will usually outperform any other auction design. Recall that the "Anglo-Dutch auction" is an English auction that is conducted until the number of bidders remaining in the competition is one more than the number of items that are being auctioned. At this stage, the English auction ends, and the bidders are invited to submit a single-sealed bid, which may not be lower than the last bid made in the English auction.
In the context of bankruptcy auctions, the most important feature of the Anglo-Dutch auction is that it motivates weak bidders to participate because they understand that once they get past the English auction phase, they too can compete with powerful, large bidders in a one-shot round, during which each bidder submits her highest bid in a closed envelope. Indeed, in a simple English auction, strong bidders can always offer one dollar more than weak bidders. By doing so, strong bidders can secure the winning bid. Because the participation of even one more bidder is critical and may tilt the auction from a failure towards success, the Anglo-Dutch design encourages participation in a manner that effectively brings even "small" bidders to the auction house.
Moreover, there is a secondary benefit to employing an Anglo-Dutch auction. An Anglo-Dutch auction undermines cartelistic behavior because the sealed part of the auction generates an incentive for the members of a cartel to defect (i.e., submit a higher bid and win the auction alone). Such an outcome is harder to obtain otherwise because an open auction allows the members of the cartel to monitor one another constantly and react immediately to any defections.
b. The Veto Feature
Consider next the veto feature of the suggested auction design. We argue that the secured creditor should be given a right to veto the sale to prevent any possible cartelistic behavior and to protect her interests against extremely low bids.
One could argue that a reserve price, perhaps even one set by the secured creditor herself, may be superior to the suggested veto mechanism. The reserve price could assure the secured creditor that the auction will not end with a sale of the encumbered assets at a price that does not reflect their market value. However, setting a correct reserve price is a complex task in the context of a bankruptcy auction. Doing so would force the secured creditor to commit in advance to a certain market value of the auctioned assets, which could later induce either inefficient sales due to an excessively low reserve price or the loss of efficient sales because of a reserve price that was set too high. Moreover, incorporating a reserve price into the auction mechanism demands an answer to several questions concerning, for example, the disclosure of this price in advance to the bidders. It is also unclear how a reserve price would impact the intensity of the competition among bidders. Thus, a veto mechanism should be expected to outperform a reserve price mechanism. Indeed, as the auction unfolds, the secured creditor, equipped with an entitlement to later veto the results of the auction, can draw various signals regarding the value of the auctioned assets. The secured creditor can do so by observing various factors, including the manner in which the auction is conducted, the identities of the participants, their behavior, and the intensity of the bidding competition.
Nevertheless, because a veto right may have a considerable "chilling effect" on potential bidders' participation-as each of them may fear that upon the auction's conclusion and after spending a substantial amount of money on preparing for the auction, the secured creditor will evoke her veto and annul the sale-the right to execute this veto should be conditioned on the secured creditor's payment of the costs of the winning bidder. The amount of the costs should be pre-determined such that all agents understand what is at stake. Additionally, the amount should include at least the cost of conducting an evaluation of the auctioned assets (which is usually the most expensive cost associated with participating in the auction). Imposing a price on evoking the veto guarantees that the secured creditor will not evoke her veto right unless she finds it to be absolutely necessary. Moreover, having a veto that is conditioned on the payment of costs to the winning bidder encourages the bidders to submit higher bids to win the auction. The bidders understand that, on the margin, it is better to win the auction than to lose because winning entitles them to either the auctioned asset or compensation from the secured creditor. In contrast, if they do not win the auction, they are not entitled to anything.
However, the secured creditor's right to veto the sale should expire once the bidding price exceeds the amount of the secured claim. Indeed, if the bidding price exceeds the amount of the secured claim, the secured creditor becomes disinterested in maximizing the proceeds of the sale (i.e., she is expected to be paid in full and no longer gains from any additional increase in the sale price) and should therefore not enjoy any special privileges, including an entitlement to veto the auction.
c. Other Features
As for the participation of investment bankers in the auction process, it seems that their involvement usually has a positive impact on the auction's success. This impact also justifies their success fee. Indeed, the relatively uncompetitive environment of bankruptcy auctions merits any boost possible. Investment bankers disseminate information among potential bidders and may play an important role in overcoming informational asymmetries. Therefore, it is useful to set their fee based on the results of the auction.
However, investment bankers should not be paid a success fee based on a percentage from the sales price in the cases in which a veto has been evoked. If the secured creditor decides to evoke her veto, the investment bankers should not be compensated with the full success fee. Instead, they should be paid a much smaller amount that reflects their failure to generate effective competition or obtain an adequate price.
We suggest that investment bankers should not be allowed to push for a pre-auction contract with a "stalking horse." Although the willingness of the "stalking horse" to commit to a certain price generates a positive signal to other potential bidders, the costs of contracting with such a buyer are quite high. The process of negotiating with this buyer may induce manipulation and collusion with interested individuals associated with the bankrupt firm or with the investment bank. Additionally, the special benefits extended to this buyer may generate a "chilling effect" over the competition, as the other potential bidders understand that outbidding the "stalking horse" may be difficult. At the same time, the positive signal generated by the "stalking horse's" commitment to buying the assets can also be generated by the English auction because other potential bidders can observe from up-close as one buyer (who would otherwise be the "stalking horse") is willing to bid high.
Finally, an important question concerns owner participation in the auction (as part of a possible "sale-back"). There should be no reason to prevent the incumbent owner (i.e., the incumbent shareholders of the firm) from participating in the auction as long as the owner's identity is disclosed to the other bidders. Because of the owner's informational advantage, the owner's participation in the auction will generate a positive signal to the other bidders and will therefore strengthen their confidence to bid high. Any "chilling effect" created by these bidders' fears of overbidding the owner (who is assumed to know best the true value of the auctioned assets) will be mitigated by the English auction format. However, if the bidders can only suspect owner participation and do not know for certain that the owner is participating, the positive signal generated by the owner's participation is weakened, and the "chilling effect" becomes stronger. Thus, owners who wish to participate in the auction should be forced to disclose their identities at the beginning, and the bankruptcy court should threaten non-disclosure with sanctions. Such a regime may not be sufficient in those cases in which owners are determined to conceal their identity, so lawmakers should consider defining such behavior as fraud.
B. Credit Bidding-A Critical Analysis
Auction theory can help resolve the disagreement concerning "credit bidding" as well because, at its basis, this question is a policy question.208 The legal controversy surrounding "credit bidding" originates from a recurring conflict of interests between a secured creditor holding an under-secured claim and the firm's incumbent shareholders and management. Although the secured creditor is usually interested in maximizing the proceeds of the sale, experience shows that the shareholders and managers of the bankrupt firm may be motivated by interests other than maximizing the proceeds from the sale, such as selling the auctioned assets to a particular buyer. Such a buyer may include a "white knight" or even an entity in which the shareholders of the auctioned firm have invested their capital (i.e., a "sale back").209 Secured creditors holding an undersecured claim should be protected against such opportunism.
"Credit bidding" seems to be able to protect the secured creditor against such danger. However, the auction design that we suggested in the previous section, which incorporates a conditioned veto right for the secured creditor, renders an entitlement to "credit bid" unnecessary. Indeed, we argue that our auction design is superior. Consider the following.
It has been suggested that "credit bidding" is beneficial in three respects.210 First, credit bidding may prevent "white knight" buyers (who may not even offer the highest price) from being favored. Second, "credit bidding" can also increase the number of bidders sufficiently familiar with the auctioned assets and thus push the bidding higher, as "credit bidding" brings the secured creditor- who is usually assumed to be well acquainted with the firm-to the auction house. Third, "credit bidding" reduces the costs of bidding and therefore minimizes the transaction costs. Indeed, it has been argued in this context that if secured creditors are denied the option of "credit bidding," they will not refrain from participating in the auction. Instead of "credit bidding," they will obtain a cash loan from a third party, even if only for a day, to bid with cash.
Although it is true that "credit bidding" may prevent a detrimental collusive interaction between the bankrupt firm (or its shareholders and its management) and a "white knight," it is not the only means to this end. Our suggestion of a veto right for the secured creditor can generate the same effect.
It is also true that if the secured creditor is genuinely interested in purchasing the auctioned assets, "credit bidding" is an excellent tool for reducing transaction costs. However, the relevant set of assumptions to which the designers of the auction should adhere is different. Although the secured creditor may be quite familiar with the auctioned assets, she is usually not interested in these assets per se.211 Thus, allowing the secured creditor to win the auction and redeem the auctioned assets in exchange for her claim does not promise much of anything to the creditors. Both the secured creditor and the "white knight" understand perfectly that the secured creditor is not really interested in the auctioned assets and would only like to prevent a sale from being concluded on bad terms. The only real option for a secured creditor who outbids the "white knight" in the auction by using credit is to sell the auctioned assets to that "white knight," albeit under perhaps more favorable terms, in a post-auction private transaction between the two parties. Note, however, that although such a post-auction transaction may be executed at a higher price than that bid by the "white knight" in the auction, the secured creditor should not expect too much out of such a transaction. The "white knight" will offer during the negotiations a price close to the one she bid in the auction, arguing this price to be the true value of the asset, as discovered in the recent market test (i.e., the auction). Moreover, a post-auction sale will require the secured creditor to incur more costs, particularly the costs associated with shopping the assets to different investors and managing the assets until a new transaction is concluded. Finally, the "white knight" will anticipate this result and will submit low bids to begin with (especially in uncompetitive auctions). By doing so, the "white knight" will strengthen her bargaining power in any post-auction negotiation.
At the same time, "credit bidding" may also have a detrimental chilling effect on potential bidders and the intensity of the auction competition because the secured creditor can bid "for free" up to the value of her secured claim. Other bidders may fear that the secured creditor will overuse her "credit bidding" entitlement in an attempt to drive the bidding higher. The chilling effect becomes even stronger if the potential bidders fear that the secured creditor will not incur any costs should she decide to "credit bid."
According the secured creditor a veto right, which is conditioned on the secured creditor paying the real costs of bidding (in a pre-determined amount) to the winner in the auction, restrains the secured creditor and enhances bidder participation and competition. Potential bidders will understand that the secured creditor will be able to veto the results of the auction, but because of the price attached to the veto, she will display self-restraint and refrain from vetoing strategically. These potential bidders expect the secured creditor to veto the results of the auction only if the winning bid is considerably lower than the secured creditor's valuation of these assets. Therefore, these bidders are encouraged to join the auction and bid higher. Enhancing competition in the problematic environment of bankruptcy auctions is invaluable.
Moreover, it has already been argued that "credit bidding" may threaten junior unsecured creditors in certain situations. 12 Consider the cases in which the secured creditor can both "credit bid" and control the terms of the auction, specifically the timing. In the cases in which the secured creditor may or may not be undersecured, depending on the exact time at which the auction is executed, secured creditors can employ the following pattern. They can push for a quick auction (which would be less competitive and less intense), "credit bid" to win the auctioned assets, and consequently eliminate any unsecured claims against these assets. These unsecured claims will be wiped out, but they would not have been underwater if the auction had been postponed and the auctioned assets had been "shopped around." Therefore, allowing secured creditors to insist on reorganization plans that include an entitlement to "credit bid" can undermine the rights of unsecured creditors. However, this result cannot happen if an ADVA is employed.
Thus, an auction design that protects the secured creditor with a conditioned veto (and that follows the general Anglo-Dutch design of the auction) should be superior to any other design in the context of bankruptcy auctions, including one that extends the secured creditor with an entitlement to "credit bid."
C. The Absence of a Regulator
Employing auction theory in real life is not an easy task. Similar to any social science, game theory is not an exact science when applied to real-world scenarios. Theoretical models and conjectures may collapse under the weight of a complex economic reality inhabited by players who sometimes behave irrationally. Large-scale auctions conducted by governmental entities, such as the FCC, to allocate scarce public assets reveal that auction theory is often best used within the confines of a trial-and-error process. In other words, sometimes the best contribution that auction theory can offer is to aid the design of the next auction following a failed one. For example, the FCC executed a series of carefully designed auctions of spectrum licenses during the 1990s.213 These auctions were engineered with the help of several auction theory experts, some of whom were recruited from academia.214 These experts discovered several new, previously unanticipated problems only after several auctions had already been concluded.215 For example, one problem concerned the post-auction financial distress experienced by the winning bidders, who could not pay the FCC to meet their bid obligations.216 Another problem concerned a failure to allocate licenses to the preferred sectors.217 These problems required the FCC and its experts to go back to the drawing board and redesign the format of future auctions.218
Thus, when considering the procedural contribution of auction theory to the design of bankruptcy auctions, we find that one of the main problems undermining any effort to execute efficient bankruptcy auctions-which should also alert lawmakers-concerns the absence of a regulator who can implement a structured trial-anderror procedure. For example, unlike FCC auctions, where the FCC can employ auction theory to design what it believes to be the optimal auction given the circumstances, execute the auction, closely follow the results, and redesign a better auction in the future, bankruptcy auctions currently cannot follow a similar track. The task of designing any specific bankruptcy auction is privatized, and although an accepted best practice may evolve among practitioners, this practice can hardly entertain systematic attempts to improve- whether ad hoc or permanently-the design of these auctions.
For this reason, it may be useful to adopt a legal arrangement that dictates a default auction design that may be efficient in most cases alongside an option to alter the terms of the auction following the bankruptcy court's approval. Of course, such an approval should be given only after the court is convinced that there are compelling arguments to adopt a different auction design.
To resolve the absence of a "bankruptcy auction regulator," the bankruptcy court must dive in and increase its involvement in the regulation of auction designs. Even now, under the current regime, bankruptcy courts enjoy a considerable amount of discretion when designing bankruptcy auctions. The legislature should amend the Bankruptcy Code to deepen the bankruptcy courts' involvement in this matter.
The importance embedded in the specific design of an auction, particularly large-scale auctions, has yet to capture the attention of bankruptcy lawmakers. Mistakenly understood as an issue of only marginal importance, the art of designing bankruptcy auctions has been left unregulated and entrusted to market actors, whose interests sometimes conflict. Although a bankruptcy auction should certainly attempt to maximize the proceeds for the benefit of all claimants, its terms are currently decided by the interested claimants.
This Article attempted to highlight the importance of carefully designing bankruptcy auctions and the contribution that auction theory can offer to this goal. To demonstrate the potential contribution of auction theory, this Article suggested a novel default design of a bankruptcy auction and discussed the currently controversial issue of "credit bidding." This Article also argued in favor of increased involvement by the bankruptcy courts in the design of bankruptcy auctions.
Auction theory is an evolving science, and the law of bankruptcy auctions should evolve accordingly. In fact, both disciplines can benefit from the interaction between them. This Article highlighted the unquestionable contribution of auction theory to the design of bankruptcy auctions. However, we did so intuitively because of the absence of sufficient empirical data. Once bankruptcy auctions become an issue of interest for scholars and lawmakers, studies of these auctions will undoubtedly follow in a manner that will enrich both bankruptcy law and the general theory of auctions. Therefore, bankruptcy and auction theory scholars should turn their attention to the study of auctions in bankruptcy.
1. See Peter C. Fusaro & Ross M. Miller, What Went Wrong at Enron 178 (2002) (discussing Enron's Chapter 11 bankruptcy); Douglas G. Baird & Robert k. Rasmussen, Four (or Five) Easy Lessons from Enron, 55 vand. l. Rev. 1787, 1809 (2002) (discussing the lessons from the Enron scandal including how quickly an auction was held after Enron's collapse and bankruptcy).
2. See Matthew Futterman, Dodgers Up Against Cash Crisis, Wall St. J., Apr. 29, 2011, at B8 (discussing the reasons for the financial difficulties faced by the Dodgers).
3. See Douglas G. Baird & Robert K. Rasmussen, Chapter 11 at Twilight, 56 Stan. L. Rev. 673, 685-89 (2003) (analyzing such a case).
4. For a description and discussion of these bankruptcy auctions, see Douglas G. Baird, Car Trouble, (Univ. of Chi. Law & Econ., Olin Working Paper No. 551, 2011), available at http://ssrn.com/abstract= 1833731 (arguing that the bankruptcies of Chrysler and General Motors provide a number of lessons for corporate reorganization).
5. See, e.g., Michael Bathon, Evergreen Solar Wins Approval to Sell Most of Its Assets, Bloomberg.com (Nov. 11, 2011), www.bloomberg.com/news/201111-10/evergreen-solar-wins-court-approval-to-sell-most-of-its-assets.html (reporting the bankruptcy auction of Evergreen Solar, Inc.).
6. See, e.g., Scott D. Cousins, Chapter 11 Asset Sales, 27 Del. J. corp. L. 835, 835-37 (2002) (discussing bankruptcy auctions of dot-com companies).
7. For a European perspective, see, for example, David Hahn, When Bankruptcy Meets Antitrust: The Case for Non-Cash Auctions in Concentrated Banking Markets, 11 STAN. J. L. Bus. & Fin. 28, 30 (2005) (describing the case of non-U.S. jurisdictions with concentrated markets); S. Abraham Ravid & Stefan Sundgren, The Comparative Efficiency of Small-Firms Bankruptcies: A Study of the US and Finnish Bankruptcy Codes, 27 fin. MGMT. 28, 29 (1998) (discussing the European inclination towards creditor-friendly bankruptcy regimes); Karin S. Thorbum, Bankruptcy Auctions: Costs, Debt Recovery, and Firm Survival, 58 J. FIN. ECON. 337, 342-43 (2000) (discussing the Swedish practice of a mandatory auction for all firms entering bankruptcy).
8. See Douglas G. Baird, The New Face of Chapter II, 12 Am. Bankr. INST. L. Rev. 69, 71 (2004) (discussing Chapter 11 as a forum for market sales); Baird & Rasmussen, supra note 3, at 691 ("Selling a business in Chapter 11 is no longer a last resort but an option to be exercised at any time if it is in the creditors' interest."); Stephen J. Lubben, The "New and Improved" Chapter II, 93 K.Y. L.J. 839, 849 (2005) ("Chapter 11 offers better rules for selling a firm than state law."). Of course, the real world is complex, and an auction is not always sufficient to resolve a bankruptcy case. See also Lynn M. LoPucki & Joseph W. Doherty, Bankruptcy Fire Sales, 106 MICH. L. REV. 1, 3 (2007) (presenting empirical evidence that rather than an auction, reorganization remains essential to dealing with the financial distress of large public companies).
9. For reasons of convenience, this Article focuses on corporate rather than personal bankruptcies. However, the discussion can easily be extended to the latter context as well.
10. See Paul Povel & Rajdeep Singh, Sale-Backs in Bankruptcy, 23 J.L. ECON. & org. 710, 710 (2007) (noting that "only a small fraction of bankrupt firms are actually reorganized and most firms (in particular, small firms) are sold, either as going concerns or piecemeal"). In 2002, 56% of the large Chapter 11 cases that ended included a sale. Baird & Rasmussen, supra note 3, at 675-76. See also Douglas G. Baird & Robert K. Rasmussen, The End of Bankruptcy, 55 Stan. L. Rev. 751, 751-52, 786 (2002) (noting that many firms use Chapter 11 merely to sell their assets and that ongoing concern sales in general have long been "the method of choice" for dealing with financially distressed firms); B. Espen Eckbo & Karin S. Thorburn, Bankruptcy as an Auction Process: Lessons from Sweden, 21 J. App. Corp. Fin. 38, 39 (2009) (describing the rise of an active secondary market for publicly traded distressed debt claims in the United States). For an attempt to portray a more balanced picture, see Lynn M. LoPucki, The Nature of the Bankrupt Firm: A Response to Baird and Rasmussen 's "The End of Bankruptcy," 56 Stan. L. Rev. 645, 647 (2004) (showing that "traditional" reorganizations, which do not include an auction, are also prevalent).
11. The term "reorganization" is used in this Article to describe the process of attending to the financially distressed firm's assets and changing the manner in which these assets are employed as part of the firm's course of real economic activities. Of course, not all financially distressed firms need to be reorganized to be rescued, as sometimes the source of the firm's distress lies in the right side of the balance sheet, which lists the firm's obligations, rather than the left side, which lists the assets.
12. The term "restructuring" is used in this Article to describe the process of attending to the financially distressed firm's obligations and changing its capital structure. During a restructuring process, debt may, for example, be reduced, replaced with equity entitlements, or restructured differently. Of course, if the source of the firm's hardship emanates from the left side of the balance sheet-in other words, the firm is economically non-viable-a restructuring can only temporarily alleviate the firm's financial distress.
13. See, e.g., Edith S. Hotchkiss & Robert M. Mooradian, Acquisitions as a Means of Restructuring Firms in Chapter 11,1 J. fin. intermediation 240, 24344 (1998) (finding empirical support that firms merged with bankrupt targets show significant improvements in operating performance).
14. Exceptions, most of which are found in the financial-economic literature, include: Baird, supra note 4, at 2 (discussing creditor control of the auction and credit bidding); Elazar Berkovitch, Ronen Israel, & Jaime F. Zender, Optimal Bankruptcy Law and Firm-Specific Investments, 33 eur. econ. Rev. 441, 488 ( 1997) (suggesting a "restricted auction" design, which allows creditors to prevent bankruptcy proceedings but prevents them from participating in a second-price sealed-bid auction); Sugato Bhattacharyya & Rajdeep Singh, The Resolution of Bankruptcy by Auction: Allocating the Residual Right of Design, 54 J. fin. econ. 269, 269 (1999) (discussing control over the design of the auction); Vincent S.J. Buccola & Ashley C. Keller, Credit Bidding and the Design of Bankruptcy Auctions, 18 Geo. Mason L. Rev. 99, 100 (2010) (discussing the issue of credit bidding); Edith S. Hotchkiss & Robert M. Mooradian, Auctions in Bankruptcy, 9 J. Corp. Fin. 555, 555-56 (2003) (discussing the question of mandatory versus voluntary auctions, as this feature can impact the results of the auction); Bruce A. Markell, Owners, Auctions, and Absolute Priority in Bankruptcy Reorganizations, 44 Stan. L. Rev. 69, 70 (1991) (discussing owner participation in bankruptcy reorganization plans using auction theory); Povel & Singh, supra note 10, at 711 (discussing owner participation in auctions); Alan N. Resnick, Denying Secured Creditors the Right to Credit Bid in Chapter 11 Cases and the Risk of Undervaluation, 63 hastings L.J. 323, 330 (2012) (discussing the issue of credit bidding). Of course, studies have also empirically studied how bankruptcy auctions are executed. See infra Part I.A.2 (discussing common features of bankruptcy auctions).
15. For a similar phenomenon in other contexts of corporate sales, see Jonathan R. Macy, Auction Theory, MBOs and Property Rights in Corporate Assets, 25 Wake Forest L. Rev. 85, 86 (1990) ("[T]he rich literature from the field of economics on auction models has not permeated into the legal literature.").
16. See infra Part I.B.I (arguing for the use of auctions except in contexts where bankruptcy is expected to pursue redistributive goals).
17. See infra Part I.B.I (discussing whether creditors can bid with their claim for credit in a bankruptcy auction).
18. See infra Part III (explaining how auction theory can be used to improve bankruptcy auctions).
19. See infra Part I.A.2, III.C (arguing that auction theory helps highlight the current problems with bankruptcy auction proceedings).
20. See infra Part III.A.2 (detailing the ADVA design for auctions).
21. An Anglo-Dutch auction begins by announcing a low amount for the price and the amount increases as the bidders continue to raise it, until the number of bidders remaining in the competition is one more than the number of items that are being auctioned. At this stage, the bidders are invited to submit a single sealed bid, which may not be lower than the last bid made thus far. In a sealed-bid auction, each bidder submits a bid in a sealed envelope; all the envelopes are opened at the same time; and the bidder who has submitted the highest bid wins the auction. An English auction begins by announcing a low amount for the price and the amount increases as the bidders continue to raise it, until only one bidder remains, with that bidder being the winner of the auction. See infra Part II.? (discussing the four basic types of bankruptcy auctions).
22. See infra Part III.? (arguing that ADVA is a superior auction design and that credit bidding is unnecessary).
23. See infra Part III.C (arguing that the absence of a regulator makes it difficult to hold efficient bankruptcy auctions).
24. See Robert G. Hansen, Auctions of Companies, 39 ECON. INQUIRY 30, 30-34 (2001) (describing how non-bankruptcy auctions have become standardized).
25. A formal bankruptcy procedure may be initiated by either the firm (voluntary case) or its claimants (involuntary case). 11 U.S.C. Â§Â§ 301(a), 303(a) (2006).
26. Id. Â§Â§ 701-784.
27. Id. Â§Â§ 1101-1174.
28. See id. Â§ 704(a)(1) (stating that the trustee shall "collect and reduce to money the property of the estate"); William C. Whitford, What's Right About Chapter II, 72 Wash. U. L.Q. 1379, 1402 (1994) (indicating that appointing a trustee in Chapter 11 should also trigger an auction solution).
29. 11 U.S.C. Â§ 363(b)(1).
30. Id. Â§ 103(a).
31. See Jason Brege, An Efficiency Model of Section 363(b) Sales, 92 Va. L. Rev. 1639, 1640 (2006) (questioning why the Bankruptcy Code allows failing businesses to "escape the rigor" of bankruptcy plan confirmation using Section 363(b)).
32. See id. at 1643 (discussing how Section 363(b) allows debtors-inpossession to wastefully or improperly dispose of assets early in the bankruptcy process before creditors have complete information regarding reorganization).
33. See id. ("Section 363(b) appears to offer a side door to escape the rigors of the typical bankruptcy plan confirmation.").
34. See id. at 1640 (discussing the varying "business justifications" courts have accepted as warranting a Â§ 363(b) sale).
35. See In re Lionel Corp., 722 F.2d 1063, 1070-72 (2d Cir. 1983) (stating that, with a liberal reading of Â§ 363(b), a bankruptcy judge has considerable discretion to approve a sale, but also must articulate sound business justifications for his decision, and cannot approve a sale on the basis that the company's creditors demanded it).
36. Id. at 1071.
38. See Brege, supra note 31, at 1653 (quoting the relevant case law).
39. See id. at 1640-42 (discussing the prevalence of these sales since the 2000s); LoPucki & Doherty, supra note 8, at 12-14 (describing how the rate of Â§ 363 sales increased during the 1990s).
40. See Lynn M. LoPucki, Courting Failure: How Competition for Big Cases Is Corrupting the Bankruptcy Courts 168-71 (2005) (attempting to explain this phenomenon by referring to the intense competition over the big cases among the courts).
41. The reasons for the debtor's objection to the secured creditor bidding with its claim have not yet been analyzed by the courts that have had to make a decision on this issue. Obviously, the debtor (i.e., the debtor-in-possession's management) and the secured creditor disagree on the best way to dispose of the debtor's assets. A more detailed analysis of this conflict will be provided in Part I1I.B.
42. 11 U.S.C. Â§ 363(k) (2006) (specifying that in an auction "of property that is subject to a lien that secures an allowed claim, unless the court for cause orders otherwise the holder of such claim may bid at such sale, and, if the holder of such claim purchases such property, such holder may offset such claim against the purchase price of such property").
44. Id. Â§ 1129(a)(8).
45. See, e.g., In re River Road Hotel Partners, LLC, 651 F.3d 642, 645 (7th Cir. 2011) (showing how lenders filed objections to plans that would "impair debtors' interests" and "sought to sell encumbered assets free and clear of liens" without allowing lenders to credit bid).
47. See Kham & Nate's Shoes No. 2, Inc. v. First Bank, 908 F.2d 1351, 1359 (7th Cir. 1990) (citing 11 U.S.C. Â§ 1129(b)(1), which allows for a "fair and equitable" plan to be "crammed down" as described above, as justification for such a sale).
48. 11 U.S.C. Â§ 1129(b)(1).
49. Id. Â§ 1129(b)(2)(A).
50. To decide these cases, the courts must decide "whether subsection (iii) can be used to confirm every type of reorganization plan or only those plans that fall outside the scope of Subsections (i) and (ii)." Additionally, the courts must determine what "indubitable equivalent" means. In re River Road Hotel Partners, LLC, 651 F.3d at 648 (Cudahy, J.).
51. In more technical terms, the question is whether 11 U.S.C. Â§ 1129(b)(2)(A)(ii) is the exclusive method through which a debtor can cram down a plan calling for the auction of encumbered assets free of liens on a secured creditor or whether Â§ 1129(b)(2)(A)(iii) also authorizes debtors to confirm such a plan. See In re Phila. Newspapers, LLC, 599 F.3d 298, 319 (3d Cir. 2010) (Ambro, J., dissenting) (stating that 11 U.S.C. Â§ 1129(b)(2)(A) requires "cramdown plan sales free of liens" to fall under Â§ 1129(b)(2)(A)(ii), not Â§ 1129(b)(2)( A)(iii), in light of the entire Bankruptcy Code, legislative history, and drafters' comments).
52. In re Phila. Newspapers, LLC, 599 F.3d at 319; In re Pacific Lumber Co., 584 F.3d 229, 245 (5th Cir. 2009).
53. In re River Road Hotel Partners, LLC, 651 F.3d at 648-51 (consolidating two cases and currently pending before the United States Supreme Court).
54. 11 U.S.C. Â§ 1129(b)(2)(A).
55. In re River Road Hotel Partners, LLC, 651 F.3d at 649-50.
56. Id. at 650.
60. Id. at 651 n.6.
61. For example, the firm can privately negotiate a deal with its creditors. Such a deal is often known as a "workout." See, e.g., Alan Schwartz, Bankruptcy Workouts and Debt Contracts, 36 J.L. & econ. 595, 602-03 (1993) (discussing the non-bankruptcy alternative of a workout). See also generally Takashi Shibata & Yuan Tian, Reorganization Strategies and Securities Valuation Under Asymmetric Information, 19 Int'l Rev. econ. & FIN. 412 (2010) (comparing the reorganization strategies of a Chapter 11 bankruptcy and a private workout). A workout is less effective if the claimants, particularly unsecured creditors, are dispersed. See Robert Gertner & David Scharfstein, A Theory of Workouts and the Effects of Reorganization Law, 46 J. FIN. 1189, 1191 (1991) (discussing a holdout problem in workouts).
62. 11 U.S.C. Â§ 362 (2006).
63. Thus, the assets of the firm can be sold piecemeal or as a whole. The assets may be owned by individuals (the asset being the object of the sale) or by a corporate entity (the equity of which is the object of the sale).
64. See, e.g., Douglas G. Baird, Bankruptcy's Uncontested Axioms, 108 YALE L.J. 573, 579, 593 (1998) (reviewing the debate regarding the purpose of a bankruptcy proceeding and the role of the judge); Baird & Rasmussen, supra note 3, at 685-99 (describing the various redeployment alternatives); LoPucki & Doherty, supra note 8, at 5-6 (same).
65. For a detailed discussion of the legal framework, see supra Part I.A (noting that a Chapter 11 case can be converted into a Chapter 7 case if the party can be a debtor under Chapter 7's provisions). Note that an arrangement among claimants is usually associated with a Chapter 11 proceeding, but a Chapter 11 case can be converted into a Chapter 7 liquidation. 11 U.S.C. Â§ 1112.
66. See 11 U.S.C. Â§ 1126(a), (c) (requiring that the plan be accepted in each class of claims by the creditors "that hold at least two-thirds in amount and more than one-half in number of the allowed claims of such class"). Those claimants who oppose the arrangement enjoy certain protections. 11 U.S.C. Â§ 1129(a)(8), (b).
67. See 11 U.S.C. Â§ 1128(a) ("[A]fter notice, the court shall hold a hearing on confirmation of a plan"). Notice the difference between these arrangements and out-of-court arrangements (see supra note 59 and accompanying text), which do not necessarily require either supermajority acceptance by the claimants or confirmation by the court.
68. In the corporate world, the wealth of the firm's "residual owner" is affected on the margin by the decisions related to the firm's assets. A "good" decision on how to employ the firm's assets would increase that person's wealth and a "bad" decision would decrease it. See, e.g., Lynn M. LoPucki, The Myth of the Residual Owner: An Empirical Study, 82 WASH. U. L. Q. 1341, 1343-44 (2004).
69. See LucÃan A. Bebchuk, A New Approach to Corporate Reorganizations, 101 Harv. L. Rev. 775, 778-81 (1988) (suggesting a scheme to "locate" the "true" residual owner of any firm suffering financial hardship).
70. A line of the bankruptcy literature suggests alternatives to the common auction design. These alternatives are less relevant to this Article because they focus on solving the problem of evaluating the bankrupt firm's assets rather than on designing an optimal cash auction. For example, it has been suggested that a certain percentage of the firm's equity should be auctioned to ascertain the true value of the assets. See generally Mark J. Roe, Bankruptcy and Debt: A New Model for Corporate Reorganization, 83 colum. L. Rev. 527 (1983) (discussing the desirability of holding an auction to help determine the firm's value). See also Barry E. Adler, Financial and Political Theories of American Corporate Bankruptcy, 45 Stan. L. Rev. 311, 323-33 (1993) (proposing a "Chameleon Equity" scheme, where a default on obligations extinguishes the common equity of the firm and transforms the most junior creditors into equity holders); Francesca Cornelli & Leonardo Felli, Ex-Ante Efficiency of Bankruptcy Procedures, 41 eur. econ. rev. 475, 478 (1997) (proposing to auction only a control stake of the firm: 50% of the shares plus one); Oliver Hart, Rafael La Porta Drago, Florencio Lopezde-Silanez & John Moore, A New Bankruptcy Procedure That Uses Multiple Auctions, 41 eur. econ. Rev. 461 (1997) (proposing a multiple auctions procedure); Francesca Cornelli & Leonardo Felli, How to Sell a (Bankrupt) Company (London Bus. Sch. and London Sch. of Econ., Working Paper, 2000), available at http://faculty.london.edu/fcornelli/controlll.pdf (arguing against auctioning the entire ownership of the firm and for retaining an equity stake for the creditors).
71. See, e.g., Thomas H. Jackson, The Logic and Limits of Bankruptcy Law 219 (1986) (arguing that well-developed U.S. capital markets render an auction a good redeployment option); Philippe Aghion, Oliver Hart & John Moore, The Economics of Bankruptcy Reform, 8 J.L. econ. & org. 523, 526-28 (1992) (discussing suggestions to make auctions mandatory in corporate bankruptcy); Douglas G. Baird, The Uneasy Case for Corporate Reorganizations, 15 J. LEGAL Stud. 127, 128 (1986) (arguing that redeployment by the market is superior to appraisal by the bankruptcy court); Michael C. Jensen, Corporate Control and the Politics of Finance, 4 J. APP. Corp. Fin. 13, 31-32 ( 1991 ) (arguing that auction by third parties would incentivize fair pricing and, if conducted immediately, provide the most accurate information about the assets); William H. Meckling, Financial Markets, Default, and Bankruptcy: The Role of the State, 41 LAW & CONTEMP. PROBS. 13, 38 (1977) (arguing for auctions as a market-oriented approach). See also Douglas G. Baird, Revisiting Auctions in Chapter II, 36 J.L. & ECON. 633, 648-53 (1993) (discussing the possible problems with mandatory auctions); LoPucki & Doherty, supra note 8, at 8 (describing how auctions became commonplace during the 1980s).
72. See Bhattacharyya & Singh, supra note 14, at 270 (stating that auction and market-based procedures separate these two issues).
73. See Eckbo & Thorburn, supra note 10, at 44 (providing empirical evidence regarding that effect).
74. See, e.g., William T. Bodoh, John W. Kennedy & Joseph P. Mulligan, The Parameters of the Non-Plan Liquidating Chapter Eleven: Refining the Lionel Standard, 9 Bankr. Dev. J. 1, 8-9 (1992) (emphasizing the bankruptcy costs spared as an auction is executed during a Chapter 11 proceeding); Eckbo & Thorburn, supra note 10, at 40 (noting that the auctions in the Swedish system are concluded, on average, within two months, which also reduces the need to obtain interim financing for the financially distressed firm), 44^15 (reporting that bankruptcy auctions incur lower direct costs than reorganizations do).
75. See Eckbo & Thorburn, supra note 10, at 51 (arguing that "auctions are competitive when judged by the evidence on takeovers more generally, and there is no evidence that auction premiums tend to undercut the bankrupt firm's fundamental going concern value").
76. Id. at 41. In an LBO, the buyer raises money from a third party through a shell company and uses the acquired assets to repay the loan.
77. See also Fed. R. Bankr. P. 2013-14, 6004-05 (establishing rules for holding bankruptcy auctions).
78. Empirical research has demonstrated the potential contributions of investment bankers, especially those who may be considered "top-tier" bankers, to planning and executing profitable non-bankruptcy mergers and acquisitions. Their success has mostly been attributed to their abilities to identify and structure mergers with higher synergy gains and to execute these transactions promptly. For a discussion and review of the literature, see Andrey Golubov, Dimitris Petmezas & Nickolaos G. Travlos, When It Pays to Pay Your Investment Banker: New Evidence on the Role of Financial Advisors in M&As, 67 j. Fin. 271, 273 (2012) (stating that bidding firms gain more when employing investment banks for the auction process, particularly if the bank is a top-tier advisor).
79. See LoPucki & Doherty, supra note 8, at 34-35 (stating that investment bankers hired by debtors perform the listed functions).
80. See Paul Povel & Rasjdeep Singh, Stapled Finance, 65 J. Fin. 927, 927 (2010) (discussing the advantages of using stapled finance in bidding contests).
81. See LoPucki & Doherty, supra note 8, at 38 Î·. 158 (finding that the breakup fees (i.e., the payment made to a stalking horse if one is outbid in an auction; this payment is usually explained as a reimbursement) in a sample of large, publicly traded firms auctioned in bankruptcy averaged $5 million).
82. Prospective buyers can rely on the stalking horse's examination of the assets to avoid conducting their own due diligence and to appraise their own bids. Id.
83. Break-up fees are assumed to include compensation mainly for two elements: the out-of-pocket expenses that are incurred by the "stalking horse" as it prepares to execute the transaction and the risk associated with the transaction falling through (e.g., the "horse's" opportunity costs). Id. at 41.
84. For a discussion of the break-up fees in bankruptcy auctions and the manner in which courts scrutinize these fees, see Oscar Garza, Jesse S. Finlayson & Solmaz Hamidian, Rethinking the Scope of the O'Brien Decision: Why the Third Circuit's Administrative Claims Analysis Should Not Be Applied to the Debtor's Request for Approval of a Breakup Fee in Connection with Bankruptcy Sales in Chapter II Cases, 28 Cal. Bankr. J. 1 (2005) (discussing the break-up fees in bankruptcy); Mark F. Hebbeln, The Economic Case for Judicial Deference to Break-up Fee Agreements in Bankruptcy, 13 bankr. Dev. J. 475 (1997) (making the economic case for judicial deference to break-up fees); Bruce A. Markell, The Case Against Breakup Fees in Bankruptcy, 66 Am. BANKR. L.J. 349 (1992) (arguing against break-up fees in bankruptcy); Monica E. White, Give Me a Breakup Fee: In re Reliant Energy Channelview LP and the Third Circuit's Improper Rejection of a Bankruptcy Bid Protection Provision fin re Reliant Energy Channelview LP, 594 F.3d 200, 2010), 48 Hous. L. Rev. 659 (2011) (discussing break-up fees in light of a negative circuit court treatment).
85. See, e.g., In re Reliant Energy Channelview LP, 594 F.3d 200, 203 (3d Cir. 2010) (setting a $5 million overbid requirement). See also C.R. Bowles & John Egan, The Sale of the Century or a Fraud on Creditors?: The Fiduciary Duty of Trustees and Debtors in Possession Relating to the "Sale" of a Debtor's Assets in Bankruptcy, 28 U. Mem. L. Rev. 781, 808-09 (1998) (discussing mechanisms such as "topping fees," "overbid protections," "rights of first refusal," "prepayment of due diligence expenses," and reimbursement of due diligence expenses under Â§ 503(b), which "either compensate the party bidding on the property (stalking horse) if its bid fails or make it more expensive for a competing offer to prevail over the stalking horse's bid or both;" the second category consists of "control" incentives, such as "lock-up agreements," "window shop," and "no shop" clauses, "all of which greatly inhibit third-party bidders from making compelling offers on the property by granting the stalking horse a significant degree of control over the debtor's assets, operations, or both").
86. See LoPucki & Doherty, supra note 8, at 41-42 (finding that second bidders appeared in only 35% of the cases in which a stalking horse participated in the auction; stating that second bidders outbid the latter in only 17% of the cases).
87. The legal framework in which the court can intervene is usually that of administrative expenses. If these expenses are needed to preserve the bankruptcy estate, they receive priority on other pre-bankruptcy claims. See 11 U.S.C. Â§ 503(b) (2006) (outlining what qualifies as administrative expenses). The court may approve or disapprove of paying break-up fees as part of these expenses. See, e.g., Calpine Corp. v. O'Brien Envt'l Energy, Inc. (In re O'Brien Envt'l Energy, Inc.), 181 F.3d 527, 537 (3d Cir. 1999) (discussing criteria involved for awarding break-up fees).
88. Calpine Corp., 181 F.3d at 537.
89. 11 U.S.C. Â§ 363(b)(1) (2006).
90. 11 U.S.C. Â§ 363(b)(1), (f). This characteristic cannot be taken lightly. In Canada, for example, the debtor is less capable of executing a sale that disposes of past liabilities. Stephanie Ben-Ishai & Stephen J. Lubben, Sales or Plans: A Comparative Account of the "New" Corporate Reorganizations, 56 McGlLL L.J. 591,594(2011).
91. Bankruptcy costs consist of direct costs, which include mostly administrative costs and fees (such as the fees paid to the professionals advising the firm), and indirect costs, which mostly include the opportunity costs associated with having to operate the firm under the constraints of the bankruptcy proceeding (such as the cost of having to obtain court approval for every major business decision). See, e.g., Ben Branch, The Costs of Bankruptcy-A Review, 11 Int'l Rev. Fin. Anal. 39 (2002) (discussing the various costs); Stephen P. Ferris & Robert M. Lawless, The Expenses of Financial Distress: The Direct Costs of Chapter 11, 61 U. Pitt. L. Rev. 629 (2000) (same).
92. See, e.g., Bank of Am. Nat'l. Trust & Sav. Ass'n. v. 203 N. LaSalle St. P'ship, 526 U.S. 434, 457 (1999) (noting that the best way to determine value is exposure to a market); Hotchkiss & Mooradian, supra note 13, at 244 (finding empirical support for the efficient redeployment of assets in bankruptcy auctions).
93. See, e.g., Bhattacharyya & Singh, supra note 14, at 270-71 (emphasizing the use of the auction as a method for preventing costly negotiations over the distribution of the proceeds).
94. See, e.g., Eckbo & Thorburn, supra note 10, at 40-41 (reporting that this result also occurs in the United States once an auction is chosen).
95. See, e.g., id. at 40 (discussing the tendencies of financially distressed firms' managers to prefer reorganization, which would leave them at the company's helm and allow them to continue to enjoy "private control benefits"). The authors also report that in the Swedish mandatory auction system, only highquality CEOs retain their job following the bankruptcy auction. Id. at 41.
96. See B. Espen Eckbo & Karin S. Thorburn, Control Benefits and CEO Discipline in Automatic Bankruptcy Auctions, 69 J. FIN. ECON. 227, 255 (2003) (showing that the operating profitability of firms auctioned as going concerns is on par with that of industry competitors for several years; stating that less than 20% report operating losses).
97. For a discussion of the pros and cons of attempting to achieve redistributive goals in bankruptcy, see, for example, Douglas G. Baird, Loss Distribution, Forum Shopping, and Bankruptcy: A Reply to Warren, 54 u. Chi. L. Rev. 815, 882 (1987) (arguing that redistribution rules in bankruptcy should do the same as rules that flow from other business failure); Elizabeth Warren, Bankruptcy Policy, 54 u. chi. l. Rev. 775, 789-90 (1987) (arguing that redistribution of costs of a bankrupt estate may result in a necessary loss to creditors, better enabling the bankrupt firm to survive).
98. See Aghion, Hart & Moore, supra note 71, at 527 (arguing that few bidders will join the auction because of the costs of preparing a bid and the risk of losing those resources if the bid is not accepted); LoPucki & Doherty, supra note 8, at 35 (finding that an average of 1.6 bidders exist per auction in a study and that in 58% of the cases, only one bidder exists).
99. See Aghion, Hart & Moore, supra note 71, at 527 (highlighting the difficulty of assembling funds in the markets for high cost assets); LoPucki & Doherty, supra note 8, at 9-10 (reviewing the effect of this observation on the bankruptcy literature, where scholars have attempted to devise alternative solutions).
100. See LoPucki & Doherty, supra note 8, at 28-29 (stating that the debtors' competitors are the most likely to purchase the company).
101. See Andrei Shleifer & Robert W. Vishny, Liquidation Values and Debt Capacity: A Market Equilibrium Approach, 47 J. FIN. 1343, 1358 (1992) (suggesting that when a firm in financial disputes needs to sell assets, its industry peers are likely to be experiencing problems themselves, leading to abrupt sales at prices below value in best use). Cf. ?. Espen Eckbo & Karin S. Thorburn, Automatic Bankruptcy Auctions and Fire-Sales, 89 J. FIN. econ. 404, 421 (2008) (reporting fire-sale discounts only in the auctions that lead to piecemeal liquidation); Eckbo & Thorburn, supra note 10, at 48 (reporting findings that suggest that auctions do not systematically produce low prices if the firm is sold as an ongoing concern; however, the prices do tend to be somewhat low for piecemeal liquidations); LoPucki & Doherty, supra note 8, at 29 (finding that when industry distress is high, the recovery ratios from auctions are also high).
102. See Harvey R. Miller & Shai Y. Waisman, Is Chapter 11 Bankrupt?, 47 B.c. L. Rev. 129, 173 (2005). a creditor with a secured claim of $1 million has no incentive to assume the risk of a delayed sale under the following conditions: 1) the firm's assets can be sold immediately in a fire sale to a buyer who appears before the bankruptcy forum and offers to pay $1 million but sets a deadline of two weeks for his or her offer to be accepted or rejected; or 2) if the assets can be shopped around for a while in a manner that will produce either $1.2 million or $0.8 million in a future sale at equal probabilities.
103. See Douglas G. Baird & Robert K. Rasmussen, Control Rights, Priority Rights, and the Conceptual Foundations of Corporate Reorganizations, 87 Va. L. Rev. 921, 957-58 (2001) (explaining that "in many workouts outside of bankruptcy, a senior secured lender insists that, in exchange for the restructuring of its debt, the debtor promises not to oppose a motion to lift the automatic stay in a subsequent bankruptcy proceeding"); Baird & Rasmussen, supra note 10, at 78485 (noting that large firms entering Chapter 11 bankruptcy lack sufficient cash flow and are subject to lenders' controls in order to obtain additional financing).
104. See Edith S. Hotchkiss, Kose John, Robert M. Mooradian & Karin S. Thorburn, Bankruptcy and the Resolution of Financial Distress, in handbook of Corporate Finance: Empirical Corporate Finance 235, 246^7 (B. Espen Eckbo ed., 2008) (discussing empirical evidence suggesting that creditors may force a premature sale of assets); George W. Kuney, Hijacking Chapter 11, 21 Emory Bankr. Dev. J. 19, 69-70 (2004) (highlighting problems caused by change of control provisions and lender control over a Chapter 11 case); Sarah Pei Woo, Regulatory Bankruptcy: How Bank Regulation Causes Fire Sales, 99 geo. L.J. 1615, 1617 (2011) (explaining that "[w]hen banks ... are driven by financial regulatory policy to overly prefer the liquidation of their own borrowers during dowturns, the end result is often fire sales ....").
105. See Todd C. Pulvino, Do Asset Fire Sales Exist? An Empirical Investigation of Commercial Aircraft Transactions, 53 J. fin. 939, 973 (1998) (concluding that immediate cash liquidation of insolvent firms may result in failure to maximize proceeds to claimholders); Per Stromberg, Conflicts of Interest and Market Illiquidity in Bankruptcy Auctions: Theory and Tests, 55 J. Fin. 2641, 2643 (2000) (arguing that a cash auction "suffer[s] from considerable inefficiencies when the market is illiquid").
106. See LoPucki & Doherty, supra note 8, at 12, 32-37 (describing the various benefits of sale to bankruptcy practitioners).
107. Id. at 35.
108. See id. at 41-42 (noting that one technique for manipulating the auction in favor of a preferred bidder is to make him or her a stalking horse).
109. Kuney, supra note 104, at 109; LoPucki & Doherty, supra note 8, at 12, 32-34 (finding empirical support for these phenomena).
110. LoPucki & Doherty, supra note 8, at 4, 12-13.
111. A possible explanation for this buyer's haste can be found in the problem of "sale-backs." See infra note 118 and accompanying text (defining sale-backs as union shareholders and managers appearing at the auction house disguised or hidden as bidders).
112. See Eckbo & Thorburn, supra note 10, at 48 ("[I]ndustry-wide distress appears simultaneously to increase the incidence of piecemeal liquidation and to reduce prices somewhat.").
113. See Hotchkiss & Mooradian, supra note 13, at 243 (finding empirical support for the argument that asymmetric information deters bidding from potentially less informed buyers).
114. See, e.g., In re Lionel Corp., 722 F.2d 1063, 1065 (2d Cir. 1983) (showing, as an example, that one court's encouragement of additional negotiations between the parties raised the price of an auction by $7 million); Robert G. Hansen & Randall S. Thomas, Auctions in Bankruptcy: Theoretical Analysis and Practical Guidance, 18 int'l Rev. L. & eton. 159, 160 (1998) (discussing the case of FNN Inc.).
115. See B. Espen Eckbo & Karin S. Thorburn, Creditor Financing and Overbidding in Bankruptcy Auctions: Theory and Tests, 15 J. CORP. FIN. 10, 11 (2009) (describing this phenomenon in Sweden).
116. See id. at 11-12 (finding empirical support in a study conducted in Sweden).
117. In re River Road Hotel Partners, LLC, 651 F.3d 642, 651 (7th Cir. 2011 ) (Cudahy, J.).
118. See Eckbo & Thorburn, supra note 10, at 11 (documenting creditormanagement coalitions that bid in bankruptcy auctions); Povel & Singh, supra note 10, at 711 (suggesting that sale-backs are quite common and that they "chill" the competition in the auction); Per Stromberg & Karin S. Thorburn, An Empirical Investigation of Swedish Corporations in Liquidation Bankruptcy, in EFI Research Report 31, 33-34 (1996) (reporting the findings of a Swedish study showing that in approximately 54% of the cases where the firm was kept as an ongoing concern, the buyer was a coalition of incumbent management and creditors; explaining that the coalition was the only bidder in 76% of the cases); Stromberg, supra note 105, at 2643^44 (suggesting that sale-backs are common in bankruptcy auctions, particularly in illiquid asset markets). To disguise their identities, insiders either use a "straw man" to bid in the auction or hide behind a corporate entity that serves as the formal bidder.
119. Recall that the buyer at the auction purchases the auctioned assets without past liabilities.
120. See Hotchkiss & Mooradian, supra note 14, at 556-57 (arguing that the coalitions formed by creditors at the institution of bankruptcy proceedings serve as both sellers and potential buyers at auction, incentivizing overbidding and driving up the price of the asset being sold).
121. See Baird, Revisiting Auctions, supra note 71, at 635 (stating that inside owners of assets being auctioned as a part of bankruptcy proceedings have an incentive to withhold information from outsiders).
122. See, e.g., David A. Skeel, Jr., Markets, Courts, and the Brave New World of Bankruptcy Theory, 1993 Wis. L. Rev. 465, 478 n.43 (1993) (arguing that a mandatory auction scheme frequently results in the winning bidder paying too much for the asset being sold).
123. For a discussion, see infra Part Il.C.4(a)(ii) (defining the winner's curse as a buyer's realization that he or she purchased an asset for an amount higher than its market value, resulting in a loss transaction).
124. Hotchkiss & Mooradian, supra note 14, at 556-57. See also Jeremy I. Bulow & John B. Shoven, The Bankruptcy Decision, 9 bell J. econ. 437, 455 (1978) (modeling bankruptcy under the assumption that the bank plus equity holders have the bankruptcy decision power and act in their own joint interest without considering the outcome of the third set of claimants, the bondholders).
125. Hotchkiss & Mooradian, supra note 14, at 556-57.
126. Id. at 556.
127. See id. (discussing how over-bidding behavior may deter outside bidders).
128. See Eckbo & Thorburn, supra note 10, at 40 (suggesting that making bankruptcy auctions mandatory may solve some of the problems currently associated with these auctions, such as the tendencies of managers to refrain from auctioning firms with going concern value or the court's control over whether the firm is liquidated piecemeal or survives as a going concern).
129. See, e.g., Arturo Bris, Ivo Welch & Ning Zhu, The Costs of Bankruptcy: Chapter 7 Liquidation Versus Chapter 11 Reorganization, 66 J. FlN. 1253, 1301 (2006) (finding that Chapter 7 liquidations appear to be no faster or cheaper in terms of direct expenses than Chapter 11 reorganization and thus have little to offer to unsecured creditors); LoPucki & Doherty, supra note 8, at 3-4, 44 (finding that in a sample of sixty large, publicly traded firms, whose bankruptcy proceedings were concluded between the years 2000-2004, reorganizations and restructurings yielded, on average, 80% to 91% of book value, whereas auctions yielded only 35% of book value). But see Eckbo & Thorburn, supra note 10, at 40 (noting that these results need to be interpreted with caution because "given management's control of the Chapter 11 restructuring process, those bankrupt companies that end up being put up for sale are likely to be those with a relatively low going-concern value-and therefore have low recovery rates per se").
130. For a comprehensive review of auction theory, see, for example, vljay Krishna, Auction Theory (2nd ed. 2010); Flavio M. Menezes & Paulo K. Monteiro, An Introduction to Auction Theory (2005); Paul Milgrom, Putting Auction Theory to Work (2004).
131. Auction theory is a part of game theory. For a general discussion of the game theory literature, see, for example, Drew Fudenberg & Jean TirÃ³le, Game Theory (1991); Martin J. Osborne & Ariel Rubinstein, A Course in Game Theory (1994).
132. See Robert D. Cairns, The Optimal Auction-A Mechanism for OptimaI Third-Degree Price Discrimination, 20 J. econ. behavior & org. 213, 213-14 (1993) ("One of the most important questions in the theory of auctions is how to design an optimal auction mechanism."); Eric. S. Maskin & John G. Riley, Optimal Auctions with Risk Averse Buyers, 52 econometrica 1473, 1473 (1984) (studying auctions in order to "maximize the expected revenue of a seller"); Roger b. Myerson, Optimal Auction Design, 6 Math. operation Res. 58, 58 (1981) (analyzing various auction designs to optimize benefits in specific seller situations); John G. Riley & William F. Samuelson, Optimal Auctions, 71 Am. econ. Rev. 381, 382 (1981) (comparing various auction models and how they maximize seller revenue). Auction theory puts a stronger emphasis on auction designs that maximize revenue for the seller and a weaker emphasis on designs that are efficient from the perspective of the entire market (in that the asset is allocated to the party that will maximize the benefit that can be derived from it). On occasion, these two goals can be mutually exclusive. For the distinction between the maximization of revenue and the maximization of societal benefit in the context of auction designs, see Krishna, supra note 130, at 5 (arguing that society's separate interest in efficiency is not always promoted by a revenue model because of imperfect information and high transaction costs).
133. Krishna, supra note 130, at 2-3; Jean-Jacques Laffont, Game Theory and Empirical Economics: The Case of Auction Data, 41 eur. econ. Rev. 1, 2 (1997).
134. For a survey of empirical experiments, see Brent R. Hichman, Timothy P. Hubbard & Yigit Saglam, Structural Econometric Methods in Auctions: A Guide to the Literature, J. ECONOMETRIC METHODS 67 (2012).
135. See, e.g., infra Part U.E. (discussing the FCC auctions for the allocation of spectrum licenses); Ken Binmore & Paul Klemperer, The Biggest Auction Ever: The Sale of the British 3G Telecom Licences, 112 econ. J. 74, 94-95 (2002) (discussing the auctions for the allocation of the third-generation mobile phone operator licenses in England).
136. See, e.g., Paul Klemperer, How (Not) to Run Auctions, 46 eur. econ. Rev. 829, 830 (2002) (describing the auctions for the allocation of spectrum licenses in Switzerland).
137. As this Article is not directed at economists but rather at lawyers and policymakers, we see no reason to cite the formulas and mathematical proofs brought for the various claims. Similar to Klemperer, we believe that if insights derived from theoretical research are not intuitively understandable, they should be approached with caution. See Paul Klemperer, Using and Abusing Economic Theory, 1 J. EUR. ECON. ASSOC. 272, 273 (2003) (affirming Alfred Marshall's notion that mathematics should be translated into plain language and applied to real life examples in order to be relevant).
138. The following discussion already makes several assumptions regarding the nature of the auction, which may be altered in either auction theory in general or the context of bankruptcy auctions in particular. We make these assumptions because bankruptcy auctions rarely relax these assumptions. For example, we assume that the bankruptcy auctions discussed in the paper are all "common-value auctions" or "affiliated-value auctions" (i.e., the auctioned asset is worth the same amount for each bidder, although each bidder may estimate this amount differently). "Common-value auctions" are opposed to "private-value auctions," in which each bidder assigns a potentially different value to the auctioned asset. See, e.g., Hal R. VarÃan, Intermediate Microeconomics-A Modern Approach 316 (8th ed. 2009).
139. Paul R. Milgrom & Robert J. Weber, A Theory of Auctions and Competitive Bidding, 50 econometrica 1089, 1090 (1982).
140. Id. at 1103-04.
141. For example, one characteristic of the Japanese auction is that it provides the bidders with exact information regarding the asset valuations given by the various bidders. This information can be valuable for the bidders, and according to economic theory, its dissemination can increase the revenue generated by the auction. For additional discussion, see id. at 1104 (explaining the variants of English auctions).
142. Id. atl089n.5.
143. Id. at 1090 n.7.
144. See William Vickrey, Counter speculation, Auctions, and Competitive Sealed Tenders, 16 J. FIN. 8, 8 (1961) (finding that, in an auction where the award price is equal to the second highest bid price rather than the highest bid price, allocation of resources is improved without being prejudicial to the interests of sellers and buyers).
145. An example would be the "all-pay auction," in which all bidders pay their bids, regardless of whether they win the auction. This type of auction is used primarily in charity events. See Michael R. Baye, Dan Kovenock & Casper G. deVries, The All Pay Auction with Complete Information, 8 ECON. THEORY 291, 291-92 (1996) (describing the "first price all-pay" auction model, where all players, even the losers, pay the auctioneers their bids); Vijay Krishna & John Morgan, An Analysis of the War of Attrition and the All-Pay Auction, 72 J. ECON. THEORY 343, 344 (1997) (defining the "war of attrition" and "all-pay" auction formats, which yield greater revenue than second- and first-price auctions). Another type is the "auction with a buy price," according to which the seller determines in advance a price that, if bid, will end the auction immediately. The bidder offering the pre-determined price will be proclaimed the winner. These auctions are used on various internet sites. See Zoltan Hidvegi, Wenli Wang & Andrew B. Whinston, Buy Price English Auction, 129 J. econ. theory 31, 32-33 (2006) (discussing the different types of buy-price auctions and the benefits of these auctions over other models); Stanley S. Reynolds & John Wooders, Auction with a Buy Price, 38 econ. theory 9, 9-10 (2009) (describing eBay and Yahoo's development of the buy-now online auction format and its popularity); Quazi Shahriar & John Wooders, An Experimental Study of Auction with a Buy Price Under Private and Common Values, 72 games & econ. behavior 558, 559 (2011) (finding that a buy-now auction has a "positive and statistically significant effect on seller revenue" and "lowers the standard deviation of revenue"); Nicholas Shunda, Auction with a Buy Price: The Case of Reference-Dependent Preferences, 67 Games & Econ. Behavior 645, 646 (2009) (discussing the author's model of auctions with temporary buy prices where bidders use reserve prices and buy prices to formulate a reference price). Another interesting auction format is the "two-stage sealed-bid auction." In the first stage of the auction, the bidders submit sealed bids. After the envelopes are opened, the two highest bidders are invited to submit another sealed bid, which may not be lower than their original bids. Perry, Wolfstetter & Zamir have shown that this method leads to the same result achieved in an English auction without incurring the disadvantages of that type of auction. See Motty Perry, Elmar Wolfstetter & Shmuel Zamir, A Sealed-Bid Auction That Matches the English Auction, 33 Games & econ. behavior 265, 266 (2000) (comparing English auctions and two-stage sealed-bid auctions).
146. The proposal was formulated for the British auctions of spectrum licenses, although it was ultimately used in the Italian auctions of spectrum licenses. See Paul Klemperer, Auction: Theory and Practice 88-89, 11617, 178-83 (2004) (describing how Anglo-Dutch auctions present an alternative to ascending and sealed-bid auctions by creating a hybrid form "which often captures the best features of both").
147. See Vickrey, supra note 144, at 20 (noting that sealed-bid auctions can be analyzed the same way as a Dutch auction).
148. See id. at 20-22 (discussing the advantages and disadvantages of the second-price method).
149. For a full explanation of the term "dominant strategy" in this context, see Preston R. McAfee & John McMillan, Auctions and Bidding, 25 J. ECON. LIT. 699, 708 (1987) (discussing "dominant strategy" in relation to Engish and secondprice auctions).
150. The equivalence between these two mechanisms is only partially accurate because in common or affiliated valuation environments, the English auction has an advantage over the second-price auction in that its open process offers the bidders information regarding the other bidders' valuations. Each bidder can use this information to update his or her own valuation. For a more in-depth explanation, see Krishna, supra note 130, at 4.
151. For a general discussion of why second-price auctions are so rare, see Michael Rothkopf, Thirteen Reasons Why the Vickrey-Clarke-Groves Process Is Not Practical, 55 operational Res. 191 (2007) [hereinafter Rothkopf, Thirteen Reasons]Â·, Michael Rothkopf, Thomas Teisberg & Edward Kahn, Why Are Vickrey Auctions Rare?, 98 J. Pol. econ. 94 (1990) [hereinafter Rothkopf et al., Why Are Vickery Auctions Rare?]. The second-price auction is used in various situations including the sale of certain financial instruments. Michael H. Rothkopf & Ronald M. Harstad, Two Models of Bid-Taker Cheating in Vickrey Auctions, 68 J. bus. 257, 258 n.l (1995) [hereinafter Rothkopf & Ronald, Two Models]. It is also frequently used in stamp auctions. See Stuart e. Thiel & Glenn h. Petry, Bidding Behavior in Second-Price Auctions: Rare Stamp Sales, 1923-1937, 27 app. econ. 11,11-16 (1995) (describing how, although stamp auctions allow bids by mail and floor bidders, all bidders participate in a second-price auction by making the selling price one increment over the second-highest bid if a mail bidder has the highest bid).
152. See Rothkopf et al., Why Are Vickrey Auctions Rare?, supra note 151, at 102 (arguing that second-price auctions are rare because, inter alia, bidders are afraid of bid-takers using fraudulent price-enhancing activities); Rothkopf, Thirteen Reasons, supra note 151, at 193-94 (explaining how Vickrey auctions are susceptible to cheating by the bid taker since the seller can submit an artificial bid between the two best bids and thus take some of the winning bidder's profit).
153. See Rothkopf, Thirteen Reasons, supra note 151 at 191 (outlining various reasons for the rarity of second-price auctions); Rothkopf et al., Why Are Vickrey Auctions Rare?, supra note 151, at 103 ("it could reveal to others with whom the firm must subsequently negotiate precisely how much it can yield").
154. It should be noted that in light of the character of a bankruptcy auction and its subjection to legal proceedings, its results cannot be kept confidential.
155. See Rothkopf, Thirteen Reasons, supra note 151 at 193 (explaining how Vickrey auctions are prone to collusion); Rothkopf et al., Why Are Vickrey Auctions Rare?, supra note 151, at 103 (arguing that fears of cheating and bidders' disincentives to follow truth-revealing strategies explain the rare use of sealed second-price auctions).
156. For example, when an auction was held for the allocation of spectrum licenses in New Zealand in 1990, the second-price mechanism was used. As a consequence, in one case, a bidder who had bid 100,000 New Zealand dollars for a license paid the amount bid by the second highest bidder, which was 6 dollars. In another case, the highest bid was 5 million dollars, and the second highest bid was 5,000 dollars. The government was harshly criticized for these absurd results. Klemperer, supra note 146, at 110; John McMillan, Selling Spectrum Rights, 8 J. Econ. Perspectives 145, 148 (1994). If the government had used an English auction, the results would have been the same, but the highest valuation would not have been disclosed. In hindsight, the New Zealand government's mistake was that it had not established any reserve prices for the various licenses.
157. The winner's curse was first discussed in 1971 in Edward C. Capen, Robert V. Clapp & William M. Campbell, Competitive Bidding in High-Risk Situations, 23 J. PETROLEUM TECH. 641 (1971).
158. For additional details, see Paul Milgrom, The Economics of Competitive Bidding: A Selective Survey, in social goals and social Organization: Essays in Memory of Elisha Pazner 261 (Leonid Hurwicz, David Schmeidler & Hugo Sonnenschein eds., 1985) (surveying rules and strategies in Dutch, English, discriminatory, and Vickrey auctions).
159. For an analysis of practical methods for dealing with the winner's curse phenomenon and for calculating the amount by which a bid should be reduced to neutralize it, see Peter C. Cramton, Money out of Thin Air: The Nationwide Narrowband PCS Auction, 4 J. econ. & Mgm'T strategy 267, 279-84 (1995); McAfee & McMillan, supra note 149, at 720-21. For additional literature regarding the winner's curse, see, for example, john H. k.agel & Dan Levin, Common Value Auctions and the Winner's Curse (2001); Patrick Bajari & Ali Hortacsu, The Winner's Curse, Reserve Prices and Endogenous Entry: Empirical Insights from eBay Auctions, 34 RAND J. econ. 329 (2003); Robert G. Hansen & John R. Lott, The Winner's Curse and Public Information in Common Value Auctions: Comment, 81 am. econ. Rev. 347 (1991); Charles A. Holt & Roger Sherman, The Loser's Curse, 84 Am. Econ. Rev. 642 (1994); John H. Kagel & Dan Levin, The Winner's Curse and Public Information in Common Value Auctions: Reply, 81 Am. econ. Rev. 362 (1991); John H. Kagel & Dan Levin, The Winner's Curse and Public Information in Common Value Auctions, 76 Am. econ. rev. 894 (1986); Barry Lind & Charles R. Plot, The Winner's Curse: Experiments with Buyers and with Sellers, 81 Am. econ. Rev. 335 (1991).
160. See Milgrom & Weber, supra note 139, at 1091-95 (showing that English auctions allow the bidders to see other bidders' valuations).
161. This characteristic gives the English-auction mechanism a certain advantage in comparison with the second-price auction format as well. See Krishna, supra note 130, at 87 (explaining that the availability of other bidders' prices in English auctions allows active bidders to know the price at which the departed bidders have dropped out, thus allowing "active bidders to make inferences about the information that the inactive bidders had and in this way to update their estimates of the true value").
162. See Krishna, supra note 130, at 75-77, 97-101 (explaining that the expected revenue from an English auction is at least as great as the expected revenue from a second-price auction, with some noted exceptions that result in equivalent revenues).
163. See Charles A. Holt, Competitive Bidding for Contracts Under Alternative Auction Procedures, 88 J. POL. ECON. 433, 440-43 (1980) (using a game theory analysis to determine that the expected procurement cost will be higher in a competitive auction than a sealed-bid auction if the bidders are risk-averse); Krishna, supra note 130, at 38-41 (concluding that the expected revenue from a first-price auction is greater than a second-price auction if the bidders are risk-averse); Maskin & Riley, supra note 132, at 1483 (comparing and contrasting different auctions that are designed to maximize the expected revenue of a seller faced with risk-adverse bidders); Menezes & Monteiro, supra note 130, at 32-34, 70-72 (concluding that if bidders are risk-averse, revenue is higher for a first-price auction than either a second-price or English auction).
164. For additional discussion of the cartel phenomenon, see Laffont, supra note 133, at 25-26 (breaking down the dynamics of collusion in bidding and highlighting additional sources for information regarding this issue).
165. See Paul Milgrom, Auction Theory, in Advances in economic theory: Fifth World Congress 1, 27 (Truman Bewley ed., 1986) ("[Collusion is hardest to support when secret price concessions are possible, and easiest to support when all price offers must be made publicly.").
166. For additional measures that can be taken to reduce the risk of cartel activity, see generally Chantale LaCasse, Bid Rigging and the Threat of Government Prosecution, 26 RAND J. ECON. 398 (1995) (discussing the possible solutions presented by enforcement of laws against bid-rigging and prosecution for bid-rigging).
167. See Peter Cramton & Jesse A. Schwartz, Collusive Bidding: Lessons from the FCC Spectrum Auctions, 17 J. Reg. econ. 229, 241 (2000) (identifying transparency and accessibility of information about bidder identities as benefits of reporting bidder identities); Klemperer, supra note 146, at 114 (explaining that sealed-bid auctions make signaling and retaliation harder than in ascending auctions).
168. See Robert C. Marshall & Leslie M. Marx, Bidder Collusion, 133 J. econ. Theory 374, 407 (2007) (analyzing the profitability of collusion at firstprice versus second-price auctions, and concluding that collusion is more difficult at first-price (sealed-bid) auctions); Marc S. Robinson, Collusion and the Choice of Auction, 16 RAND J. econ. 141, 144 (1985) (explaining that cartels are less stable when sealed high-bid auctions are utilized because there is an incentive to defect from the cartel); Jeroen Hinloopen & Sander Onderstal, Collusion and the Choice of Auction: An Experimental Study, tinbergen institute 15 (Discussion Paper No. 10-120/1, Nov. 30, 2010), available at http://ssrn.com/abstract=1718996 (last visited Sept. 23, 2012) (arguing that collusion is expected to be more fragile in an English auction than in a first-price sealed-bid auction).
169. For example, a group of asymmetric bidders could be a market composed of small and large construction companies or of small and large software companies.
170. For example, in the FCC's 1995 auction for the allocation of spectrum licenses in Los Angeles, the strongest bidder in the region (Pacific Telephone) signaled to the other potential bidders that it intended to win the auction at any price. The auction was conducted as an English auction. Pacific Telephone even asked Paul Milgrom to explain to its potential competitors that if they acted too aggressively in making their bids, they could fall victim to the winner's curse. Many bidders were consequently deterred from competing, and Pacific Telephone won the auction at approximately one-third of the assests' market value. See Paul Klemperer, Auctions with Almost Common Values: The 'Walnut Game' and Its Applications, 42 eur. econ. Rev. 757, 761 (1998) (explaining that the price for the Los Angeles license ($26 per person) was quite low when compared to licenses in Chicago, a much smaller city than Los Angeles, where recently auctioned licenses went for as much as $31 per person); Klemperer, supra note 146, at 106- 09 (giving multiple examples when deterrence of bidders reduced competition and often dramatically diminished the seller's revenue from the auction); Milgrom, supra note 130, at 211 n.5 (explaining that the price paid per unit by Pacific Telephone was "low compared to the prices in other market areas containing such a large urban center," despite a personal rival of Pacific's CEO driving up the price for the southern California license by hundreds of millions of dollars).
171. These conclusions are appropriate for most sales environments. See Harrison Cheng, Ranking Sealed High-Bid and Open Asymmetric Auctions, 42 J. Math. Econ. 471, 473 (2006) (discussing the "Getty effect," which argues that a stronger bidder will bid higher in a sealed-bid auction than he ordinarily would in an open auction); Klemperer, supra note 146, at 21-24, 114 (giving an overview of mechanism design theory which aims to maximize the performance of an auction through rules that govern interactions between the parties, including the rules of sealed bids); Eric Maskin & John Riley, Asymmetric Auction, 67 REV. econ. studies 413, 413 (2000) (explaining that a risk-averse seller should favor a sealed high-bid auction when buyers are also risk-averse); Milgrom, supra note 130, at 149-53 (explaining the importance of auction design and that in sealed-bid auctions entry is promoted because a weaker bidder has a better chance of winning at a price a stronger bidder may have but did not bid).
172. Milgrom & Weber, supra note 139, at 1096, 1110 (showing that reporting information will raise the bidders' willingness to pay).
173. Nevertheless, even according to the theory, it is preferable for the seller to maintain confidentiality for certain types of information (e.g., the number of participating bidders in a sealed-bid auction with risk-averse bidders).
174. See Krishna, supra note 130, at 24-28 (demonstrating that expected payments in first- and second-price auctions with reserve prices of r > 0 are the same); McAfee & McMillan, supra note 149, at 713-14 (concluding that any of the four types of auctions will succeed as an "optimal selling mechanism" if the seller imposes two optimal reserve prices). It is agreed that the establishment of a reserve price does not affect the other attributes of the different types of auctions.
175. Menezes & Monteiro, supra note 130, at 25; Myerson, supra note 132, at 67; Riley & Samuelson, supra note 132, at 382-86. See also Stephanie Rosenkranz & Patrie W. Schmitz, Reserve Price in Auctions as Reference Points, 117 ECON. J. 637, 642^t7 (2007) (reaching a similar conclusion from the perspective of cognitive psychology).
176. See Jennifer Brown & John Morgan, How Much Is a Dollar Worth? Tipping Versus Equilibrium Coexistence on Competing Online Auction Sites, 117 J. POL. ECON. 668, 682 (2009) (discussing the effects of reserve prices on revenue). Regarding the argument that the optimal reserve price depends on the degree to which the seller and the bidders are risk-averse, see Audrey Hu, Steven A. Matthews & Liang Zou, Risk Aversion and Optimal Reserve Prices in Firstand Second-Price Auctions, 145 J. ECON. THEORY 1188 (2010). Barrimore and Raviv conducted an experiment in which they sold Starbucks gift certificates on eBay. An increase of the reserve price increased the price paid for the items that were sold but also led to a situation in which some of the items were not sold at all.
Thus, overall, the reserve price had a negative impact on the total revenue from the auction. Nathan Barrimore & Yaron Raviv, The Effect of Different Reserve Prices on Auction Outcomes (Robert Day Sch. Econ. & Fin. Research, Paper No. 2009-13, July 9, 2009), available at http://ssrn.com/abstract= 1432283.
177. See Milgrom & Weber, supra note 139, at 1104 (discussing the Japanese auction system, which provides bidders with precise information regarding valuations).
178. Daniel R. Vincent, Bidding Off the Wall: Why Reserve Prices May Be Kept Secret, 65 J. econ. theory 575, 576, 583 (1995).
179. The validity of this thesis is doubtful, as it can also be argued that nondisclosure of the reserve price is likely to deter risk-averse bidders from participating if they estimate that there is a reasonable chance that their valuations are lower than the reserve price.
180. See generally heuristics and biases: the psychology of intuitive judgment (Daniel Kahneman, Dale W. Griffin & Thomas Gilovich eds., Cambridge Univ. Press 2002) (discussing various documented irrational mental operations and biases); Judgment Under Uncertainty: Heuristics and Biases (Daniel Kahneman, Paul Slovic & Amos Tversky eds., Cambridge Univ. Press 1982) (discussing instances of decision makers' irrational actions).
181. Regarding the practical considerations to be weighed in the design of an auction, see Paul Klemperer, What Really Matters in Auction Design, 16 J. Econ. Persp. 169, 169-70 (2002) (discussing the need to design auctions to deter predatory behavior and collusion).
182. For a detailed analysis of the auction, its rules, and its progress, see Cramton, supra note 159 (setting forth a strategy for bidding on licenses); John McMillan, Why Auction the Spectrum?, 19 telecomm. Pol'Y 191, 191 (1995) (advocating for auctioning electromagnetic spectrum rights).
183. For example, in one case, a particular party was awarded a license in the Cape Cod area. The party sold this license to another party for $41 million. See McMillan, supra note 182, at 192 ("[Assigning licenses at random is hardly likely to put licenses into the hands of the firms most likely to use them.").
184. Omnibus Budget Reconciliation Act of 1993, Pub. L. No. 103-66, 107 Stat. 312. It is interesting to note that the objective of obtaining the maximum revenue from the auction was a secondary goal in this congressional action. The main objective was defined as the efficient allocation of spectrum frequencies. See Preston R. McAfee & John McMillan, Analyzing the Airwaves Auction, 10 J. ECON. PERSP. 159, 160, 165 (1996) ("Congress charged the FCC with encouraging an 'efficient and intensive use of the electromagnetic spectrum.'").
185. For a fascinating description of the constraints and of the auction's progress from the perspective of the FCC itself, see Evan R. Kwerel & Gregory L. Rosston, An Insider's View of FCC Spectrum Auctions, 17 J. REG. ECON. 253, 253 (2000) (discussing how economists both inside and outside of the government collaborated to design FCC auctions); Milgrom, supra note 130, at 265-79 (explaining simultaneous ascending auctions used by the FCC).
186. McMillan, supra note 156, at 151-53. John McMillan, who took part in several of the most important studies in the field of auction theory, served as the FCC's special consultant for the design of the auction and was responsible for designing the auction's rules.
187. For a detailed discussion of the issue, including the difficulties that it raised and the solutions that were found for these problems, see McMillan, supra note 156, at 153-55 (describing the benefits and consequences of the various auction forms available to the FCC).
188. Regarding the advantages of a simultaneous auction under these circumstances, see Bashkar Chakravorty, William W. Sharkey, Yosef Spiegel & Simone Wilkie, Auctioning the Airwaves: The Contest for Broadband PCS Spectrum, 4 J. ECON. & Mgm'T Strategy 345, 354 (1995).
189. Paul Milgrom and Robert Wilson, two of the most important auction theory scholars, served as consultants for two of the companies that participated in the auction. For a discussion of the considerations involved in choosing the rules for the auction, see Milgrom, supra note 130, at 1-16.
190. Furthermore, because the allocation of over 2,000 licenses was at stake, the FCC decided to allocate them into five simultaneous auctions, with the licenses divided between these auctions on the basis of the license types. For a detailed explanation of all of the considerations, see McMillan, supra note 156, at 155-57 (noting that the FCC chose from different auction forms based on the "degree of license interdependency" and the value of the license).
191. See Cramton, supra note 159, at 278 (demonstrating the extent to which the auction rules affected the auction outcome).
192. For some additional factors that the FCC was required to consider, see Chakravorty, Sharkey, Spiegel & Wilkie, supra note 188, at 356-61. These factors included the following issues: whether to divide the spectrum licenses into a few large licenses or into many small licenses; how to divide the frequencies geographically; how many electro-magnetic frequencies to allocate; whether and to what extent the incumbent cellular operators' abilities to participate in the new auction should be limited; and whether to allow several companies to group together for the purpose of joint bidding.
193. Cramton, supra note 159, at 6 Î·. 11.
194. For a stage-by-stage description and analysis of all 47 rounds, see generally Cramton, supra note 159.
195. Cramton, supra note 159, at 269. For the details of the auction, see Auction 1: Nationwide Narrowband (PCS), federal communications Commission, available at http://wireless.fcc.gov/auctions/default.htm? job=auction_summary&id=l (last visited Nov. 27, 2012) (summarizing the 1994 spectrum licenses auction). Naturally, not all of the forecasts were accurate. For example, the intention to hide the identities of the bidders by giving them secret identity numbers turned out to be inefficient and unnecessary, and the FCC decided not to follow this practice in later auctions. Id. However, these problems were minor and did not change the final conclusion regarding the auction's success.
196. For example, the FCC allocated 99 licenses in the third auction through 112 bidding rounds. The auction lasted for four months, and the FCC was paid approximately $7 billion. For details of this auction, see Auction 4: Broadband PCS A & ? Block, federal communications commission, available at http://wireless.fcc.gov/auctions/default.htm?job=auction_summary&id=4 (last visited Nov. 27, 2012) (summarizing the 1995 spectrum license auction).
197. For a critique with respect to the excessive centralization in the purchasing of licenses and the preferences given to large corporations, see generally, Gregory F. Rose & Mark Lloyd, The Failure of FCC Spectrum Auctions, Center for American Progress (May 2006), available at http://www.americanprogress.org/kf/spectrum_auctions_may06.pdf (last visited Nov. 27, 2012). Over time, it was also discovered that the bidders had engaged in several attempts at collusion. For a description of this phenomenon and the attempts to address it, see generally Cramton & Schwartz, supra note 167; Leslie M. Marx, Economics at the Federal Communications Commission, 29 Rev. indus. ORG. 349 (2006).
198. For a survey and analysis of nine auctions that allocated spectrum licenses and that were conducted in nine different countries in Europe, see generally Klemperer, supra note 136. For a detailed analysis of the planning and execution of the 2000 auction conducted in England to allocate spectrum licenses, which generated revenues of $34 billion and is considered to be the largest auction ever held, see generally Binmore & Klemperer, supra note 135. For a detailed analysis of the auction for the allocation of spectrum licenses in Germany, see Klemperer.supra note 146, at 196-205.
199. See Klemperer, supra note 146, at 119-21 (emphasizing the need to tailor auction design to the context and giving examples of auction designs that worked well in one situation but failed in others); Milgrom, supra note 130, at 247^19 ("In practice, the design of an effective auction requires a detailed knowledge of the particular circumstances in which the auction is to be run.").
200. McAfee & McMillan, supra note 184, at 172.
201. See Klemperer, supra note 181, at 169-70 (arguing that the most important features of an auction system are its robustness against collusion and its attractiveness to potential bidders).
202. Milgrom, supra note 130, at 253.
203. See supra Part I.A.I (noting that bankruptcy courts must approve any plan).
204. See supra Part I.A.2 (discussing controversy among courts regarding the meaning of Â§ 1129(b)(2)(A)).
205. See supra note 138 and accompanying text (stating that bankruptcy courts rarely relax the assumption that an auction can be characterized as a "common-value auction" and therefore the article assumes the same). See also Rothkopf et al., Why Are Vickrey Auctions Rare?, supra note 151, at 100 (mentioning work that "developed models based on an assumption that bidders had a common (but unknown) value").
206. See supra Part I.B.2 (discussing the low participation rates in bankruptcy auctions).
207. See Jeremy Bulow & Paul Klemperer, Auctions Versus Negotiations, 86 Am. econ. Rev. 180, 190 (1996) (showing that under certain assumptions, it is more profitable to sell a company through an auction with N+l bidders than through negotiation with Î bidders). See also Patrick Bajari, Robert McMillan & Steven Tadelis, Auctions Versus Negotiations in Procurement: An Empirical Analysis, 25 J. L. econ. & ORG. 372, 389 (2009) (discussing widely-held views among industry participants that competitive bidding results in lower administrative costs than negotiations do).
208. See supra Part I.A.2 (discussing the legal context of "credit bidding," which demonstrates neutral statutory language that can encompass an interpretation either mandating credit bidding or allowing debtors to proceed with a "cram down" plan denying the practice). See also Buccola & Keller, supra note 14, at 101 (noting that the question of "credit bidding" is eventually a matter left to the "sound discretion of the bankruptcy court").
209. For convenience, the following discussion shall refer to both a "white knight" and a "sale back" as a "white knight".
210. See BuceÃ³la & Keller, supra note 14, at 100 (listing three main reasons credit bidding "is a tool well-calibrated to maximize the value of a bankruptcy estate").
211. See Jared Kawalsky, The Case Against Credit Bidding: Optimal Creditor Behavior in Chapter 11 Collateral Auctions 3 (Working Paper, 2011), available at http://works.bepress.eom/jared_kawalsky/l (assuming that a denial of the secured creditor's request to "credit bid" at an auction effectively prohibits the creditor from not only bidding its credit but also from participating in auctions generally).
212. See Baird, supra note 4, at 7 n.14 (arguing that the secured creditors' right to absolute priority is an element of protecting secured creditors when those junior to them are in control).
213. See Marx, supra note 197, at 350 (noting that these auctions began in 1994). See also Charles Z. Zheng, High Bids and Broke Winners, 100 J. econ. Theory 129, 130 (2001) (highlighting one such troubled auction held in 1996).
214. Marx, supra note 197, at 350.
215. See Zheng, supra note 213, at 130-31 (discussing one such problem discovered two years after the auctions began).
216. See, e.g., id. (describing a 1996 FCC auction of radio frequencies for ten billion dollars to firms that later defaulted).
217. See Marx, supra note 197, at 352 (noting that the bidding process tends to allocate licenses inefficiently).
218. See, e.g., id. at 350 (discussing the development of new types of FCC auctions in response to the need to sell "multiple different objects at the same time").
Yaad Rotem and Omer Dekel*
* Yaad Rotem is a Visiting Scholar at the Center for the Study of Law & Society at the University of California, Berkeley, School of Law. He is also an Assistant Professor at the Center of Law & Business in Israel. Omer Dekel is also an Assistant Professor at the Center of Law & Business in Isreal. We thank Uri Benoliel and Shahar Weiler. This Article was accepted for publication prior to the Supreme Court decision in RadLAX Gateway Hotel, LLC v. Amalgamated Bank (decided May 29, 2012). This decision is therefore not discussed in the Article…