42 DUKELJ 469 (Cite as: 42 Duke L.J. 469)

Duke Law Journal

December, 1992

Peter P. Swire [FNd]

Copyright © 1992 by the Duke Law Journal; Peter P. Swire

Reproduced with Permission

TABLE OF CONTENTS

INTRODUCTION .............................................................. 471

I.  THE HISTORY AND CURRENT NATURE OF SPECIAL BANK INSOLVENCY LAW ....... 477

      A. The History of the Special Bank Insolvency Regime ................ 478

      B. The Current Superpowers of the Banking Agencies .................. 481

          1. Cross-Guarantees ............................................. 482

          2. D'Oench Powers ............................................... 483

          3. Other Special Insolvency Rules ............................... 484

      C. Summary .......................................................... 490

II.  SYSTEMIC EFFECTS OF BANK FAILURE AND SPECIAL BANK INSOLVENCY RULES .. 490

      A. Special Bank Insolvency Rules Before Deposit Insurance ........... 490

      B. Assessing Immediacy and Bank Runs as the Bases for a Special

        Insovency Regime .................................................. 494

      C. Implications of Deposit Insurance for Bank Insolvency Rules ...... 497

          1. 'Saving the Fund' as a Source of Special Rules ............... 497

              a. Reasons supporting the 'save the fund' argument .......... 498

              b. Reasons for rejecting the 'save the fund' argument ....... 501

              c. Assessing the 'save the fund' argument ................... 503

          2. FDIC Expertise as a Source of Special Rules .................. 503

      D. Summary .......................................................... 505

III.  THE PREVALENCE OF INSIDER ABUSE AND SPECIAL BANK INSOLVENCY RULES ... 505

      A. The Historical Basis for the 'Model of Insider Abuse' ............ 507

      B. The Continuing Tendency of Insider Abuse in Bank Failures ........ 510

          1. Fraud and the Nature of Banking Transactions ................. 510

          2. Information Inadequacy and the Likelihood of Insider Abuse ... 512

          3. The Relative Weakness of Bank Monitoring ..................... 513

              a. Depositor monitoring ..................................... 514

              b. Stockholder monitoring ................................... 515

              c. Monitoring by secured creditors .......................... 516

              d. Bondholder monitoring .................................... 517

       C. Implications of Insider Abuse for a Special Bank Insolvency

         Regime ............................................................ 518

IV.  PUBLIC CHOICE AND SPECIAL BANK INSOLVENCY RULES ..................... 520

      A. The FDIC and the Public Choice of Bank Insolvency ................ 521

      B. Congress and the Public Choice of Bank Insolvency ................ 525

      C. Summary .......................................................... 529

V.  APPLICATION TO CURRENT BANK INSOLVENCY PROBLEMS ..................... 530

      A. Assessing and Explaining Particular Aspects of Bank Insolvency

        Law ............................................................... 531

          1. Cross-Guarantees ............................................. 531

              a. Cross-guarantees and the 'model of insider abuse' ........ 531

              b. Cross-guarantees and systemic effects .................... 534

              c. Cross-guarantees and public choice ....................... 536

              d. Assessing cross-guarantees ............................... 537

          2. D'Oench Powers ............................................... 537

              a. D'Oench powers and systemic effects ...................... 538

              b. D'Oench powers and the 'model of insider abuse' .......... 541

              c. D'Oench powers and public choice ......................... 542

              d. Assessing D'Oench powers ................................. 543

      B. Assessing the Overall Effects of the Special Bank Insolvency

           1. An Example: Effects of Special Bank Insolvency Rules on a

         Nearly Insolvent Bank ............................................. 545

               a. Equity holders ........................................... 546

               b. Debt holders ............................................. 548

               c. Other creditors .......................................... 549

               d. Borrowers ................................................ 549

               e. Bank personnel ........................................... 550

           2. Assessing the Current Regime ................................. 551

VI.  CONCLUSION .......................................................... 555

INTRODUCTION

Even though the word "bankruptcy" derives from bank failure, [FN1] modern banks never technically go bankrupt, no matter how hard it sometimes seems they try. Instead, since well back into the nineteenth century, American banks have been subject to a special regime of bank insolvency that places their failure outside the jurisdiction of bankruptcy courts. In 1976, Robert Clark observed that "[t]he theory behind special insolvency proceedings for financial intermediaries appears not to have received careful and sustained attention." [FN2] His observation remains true today, and legal academics have almost entirely failed even to notice the existence of a special bank insolvency regime. [FN3]

For most of the past sixty years, since the creation of the Federal Deposit Insurance Corporation (FDIC) in 1933, the special bank insolvency regime probably deserved the neglect that it received. Very few banks failed, [FN4] the few failures were almost invariably quite small, [FN5] and practices developed which insulated almost everyone except shareholders and bank officers from loss in the event of failure. [FN6]

Few would think that the neglect is still deserved today. Thrift failures, which are now treated in essentially the same way as bank failures, [FN7] have risen dramatically after 1980. [FN8] The number and size of bank failures have risen too, although with a slightly later start. [FN9] By 1992, the FDIC, which manages failed banks, and the Resolution Trust Corporation (RTC), which manages failed thrifts, each had institutions totalling hundreds of billions of dollars of assets under their control and subject to the special bank insolvency rules. [FN10] Congress has responded to these failures by passing four important laws affecting bank insolvency: the Competitive Equality Banking Act of 1987 (CEBA); [FN11] the Financial Institutions Reform, Recovery, and Enforcement Act of 1989 (FIRREA); [FN12] the Crime Control Act of 1990; [FN13] and the Federal Deposit Insurance Corporation Improvement Act of 1991 (FDICIA). [FN14] Congress and the courts have provided the agencies with "superpowers" in their handling of an institution's estate. That is, the FDIC and RTC gain powers in insolvency that would not have been available to the institution pre-insolvency, or to a non-bank in insolvency-to the disadvantage of third parties. [FN15] This Article seeks both to describe reasons for the growth of special insolvency rules for banks and to make a normative assessment of the current regime.

A natural first explanation for the growth of special insolvency rules for banks would be that banking benefits from federal deposit insurance. To protect the insurance fund, the government has created a system of pervasive regulation both in and out of insolvency. I argue, however, that this explanation is seriously incomplete. Special insolvency rules for banks date back well into the nineteenth century, long before the creation of deposit insurance. In addition, the mere desire to save the insurance fund does not convincingly explain the particular ways that bank insolvency law differs from the law applying to insolvent non-bank corporations. This Article puts forth three accounts which together explain modern bank insolvency law: the systemic effects of bank runs, the increased likelihood of insider abuse in bank failures, and the political incentives of the banking agencies and Congress.

Part I of the Article traces the history of special bank insolvency law. It also briefly describes ten categories of current special rules, in order to show the broad scope of the rules examined later in the Article.

Part II shows how the systemic effects of bank failures could explain the need for special insolvency rules. Each bank failure can harm the broader economy by blocking depositors' immediate access to their funds and by creating the risk of a generalized bank run. When individual depositors perceive a risk that their bank will fail, it may be individually rational to "run" (withdraw their money from the bank). But the result of many depositors doing so would be failure of the bank, and large losses to depositors who were slow to withdraw their money. [FN16] A run benefits no one in a bank; all depositors would be as well off (if they withdrew deposits quickly) or better off (if they withdrew deposits slowly) if the run had never started. By the late nineteenth century, special insolvency rules developed to address this collective action problem: Shareholders were required to put up extra funds in the event of bank failure, liquidations were to be handled speedily, and the government was given monopoly power to close banks. [FN17] All of these measures reduced the temptation for individual depositors to run at the first sign of trouble. Amidst the bank failures of 1929 to 1933, however, policymakers decided that these measures did not work well enough. To stop the perceived systemic effects of bank failures, federal deposit insurance was extended to banks in 1933 and to thrifts the following year. [FN18]

Deposit insurance created two new reasons for special insolvency rules. Insolvency rules could be justified as necessary to save the insurance fund. In addition, because the FDIC "steps into the shoes" of the insured depositors it pays, it is the dominant creditor in almost all bank failures, and its expertise could justify special insolvency powers.

The "save the fund" and expertise arguments do not, however, explain the existence of the many rules that pick and choose just who will be disadvantaged by special insolvency rules. Part III of the Article argues that many bank insolvency rules arose from a "model of insider abuse." Under this model, it is assumed that bank and thrift failures result from fraud or egregious mismanagement rather than from the usual workings of competitive markets. Many thrift scandals of the 1980s popularized the idea that Charles Keating and other "thrift kingpins" were to blame for savings and loan losses, and thus that managers and other insiders should be made to reimburse the insurance fund for their ill-gotten gains. The case for special insolvency rules is bolstered by theoretical reasons for believing that banks might be especially susceptible to insider abuse: Banks cannot be monitored in many of the usual ways, and banking is a business that is peculiarly subject to fraud-- banks engage in numerous, liquid, and large transactions which are ready subjects of forgery and embezzlement schemes. Insiders thus may deserve to be subject to special, strict rules when their institutions become insolvent.

Parts II and III together suggest that systemic concerns and the risk of insider abuse could provide sound policy explanations for the development of special bank insolvency rules. But, as discussed in Part IV, there is good reason for fearing that political considerations, rather than sound policy, have been the source for many of the recent rules. In examining the public choice of bank insolvency rules, I argue that the FDIC's chief goal has been to maximize the bank and thrift insurance funds, while Congress's has been to prevent blame for the bailout from being traced back to itself. In the four recently passed statutes mentioned earlier, the result has been a "strict" approach to bank insolvency. Many new special rules have become law, giving the FDIC additional superpowers and redirecting a portion of the bailout cost onto the backs of bank insiders and numerous third parties. There has often been no effective counterweight to the FDIC and Congress, with the result that the recent rules are likely stricter than sound policy would warrant.

The three accounts of systemic effects, insider abuse, and public choice combine to give a good description of how the FDIC has gained its extraordinary range of special powers in bank insolvency. These major forces have all pushed toward stricter laws favoring the FDIC and hurting third parties. The result has been a trend toward greater divergence of law applying to banks before and after insolvency, and greater divergence of insolvency law applying to banks and to other corporations.

The next task, undertaken in Part V, is assessing whether the current rules are normatively desirable. Looking at two prominent areas of bank insolvency law, Part V argues that some rules have gone too far. Although solid arguments support the "cross-guarantees" power of the FDIC, which allows the agency to tax affiliated solvent banks for losses to the insurance fund, the rules derived from D'Oench, Duhme & Co. v. FDIC [FN19] have probably expanded too far, giving the FDIC and RTC undeserved victories over some legitimate borrowers from failed banks.

More generally, there are strong reasons for believing that the overall group of special insolvency rules has gone too far. Most important is what I call the "black hole" effect for nearly insolvent institutions. The many special insolvency rules make it more costly for almost anyone-investor, borrower, contractor, employee, or even depositor-to do business with a troubled bank. Once a bank becomes troubled, it is rational for any of these persons to consider a modern form of bank run, by ceasing to do business with the weak bank. The overall result of the current rules, many of which were designed to prevent bank failures, may ironically be to encourage additional failures. With so many billions of dollars of taxpayer money at stake, these potential costs of the special bank insolvency regime deserve careful attention which they have not previously received.

I. THE HISTORY AND CURRENT NATURE OF SPECIAL BANK INSOLVENCY LAW

 Little history about special rules for bank insolvency has previously been published. A striking aspect of the history is the degree to which the special regime developed before the creation of federal deposit insurance in 1933. The mere existence of deposit insurance, or of government regulation of banking more generally, thus does not sufficiently explain why the special regime has developed. The analysis in later Parts of this Article attempts to provide a more satisfactory set of explanations. This Part lays out the key portions of the history of the special bank insolvency regime and describes its current nature.

 A. The History of the Special Bank Insolvency Regime

 The perceived need for a special bank insolvency regime dates back at least to 1837 and to 1857, when Presidents Van Buren and Buchanan, respectively, introduced bills in Congress that would have provided for a federal bankruptcy system confined to banks. [FN20] Even in a period in which the federal government played no role in banking regulation, these bills, although they did not pass, show the perceived special nature of bank insolvency. [FN21]

The National Banks Act (NBA), approved in 1864 to help finance the Civil War, began the federal practice of giving bank receivers extraordinary powers. [FN22] Procedurally, the Comptroller of the Currency, rather than a court, gained the power to appoint a receiver for national banks. [FN23] Over time, doctrine developed that courts had only limited powers to interfere with actions of these "statutory receivers." [FN24] Statutory receivers also gained at least two specific superpowers. For a time, they could sue in federal courts without regard to diversity. [FN25] In addition, upon insolvency they could enforce the "double liability" provision, which allowed assessment of stockholders for up to the par value of their stock. [FN26]

A system of state-chartered banks flourished alongside the national banks regulated by the Comptroller. [FN27] Most states established special statutory regimes for supervising banks and governing insolvencies. [FN28] These regimes were established even though the states had no direct financial stake in bank failures comparable to the FDIC's deposit insurance. Most of the key elements of the NBA scheme were reproduced in the states: appointment of a statutory receiver by a government official, streamlined administration of the estate, and double liability provisions. [FN29] The general rule, however, continued to be the normal rule in bankruptcy-the bank receiver merely "stepped into the shoes" of the bank and did not, by virtue of being receiver, acquire any enhanced powers to collect debts or otherwise. [FN30]

The Great Depression and the creation of the FDIC in 1933 led to a sharp break from past practices in bank insolvency. The Bankruptcy Act of 1898 had permitted the existing Comptroller and state bank insolvency proceedings to continue, and had explicitly excluded banks from its coverage. [FN31] The FDIC became the exclusive receiver for failed national banks, and also gained the power to be receiver for state insured banks at the discretion of state authorities. Similar receivership powers were granted to the Federal Home Loan Bank Board (FHLBB) for thrifts insured by the Federal Savings and Loan Insurance Corporation (FSLIC). [FN32]

Creation of the FDIC and FSLIC began a half-century in which bank insolvency law nearly disappeared as a topic of legal interest. The number of bank and thrift failures dropped sharply after passage of deposit insurance, and most subsequent failures involved very small institutions. [FN33] The only development of continuing interest was the 1942 case D'Oench, Duhme & Co. v. FDIC, [FN34] discussed below, which has become the source of an elaborate set of rules favoring the government receiver in cases in which written bank files have not been meticulously kept.

The enormous thrift failures of the 1980s eventually caused the FSLIC to run out of money, prompting passage of FIRREA in 1989. Among its numerous provisions, [FN35] FIRREA codified the powers of the FDIC and RTC acting as "receiver" (liquidating the failed institution) or "conservator" (conserving the failed institution as a going concern in hopes of reselling it). [FN36] FIRREA continued the centralization of power over insolvencies into FDIC hands. After FIRREA, the FDIC is the exclusive receiver and conservator for failed national banks by law, and for failed state-insured banks by practice. [FN37]

The FDIC was also put on the path to controlling thrift insolvencies. FIRREA created the Resolution Trust Corporation (RTC), which serves as the receiver or conservator for all thrifts that fail between January 1, 1989 and October 1, 1993. [FN38] The FDIC was given an exclusive management role over the RTC, and the Chairman and Board of the FDIC were made the Chairman and Board of the RTC. [FN39] Furthermore, the FDIC will be the exclusive receiver or conservator for thrifts that fail after October 1, 1993.

Congress continued to expand the agency insolvency powers in the Crime Control Act of 1990 and FDICIA in 1991. Under FDICIA, the agency's powers in insolvency began to be leveraged for the first time into significant powers over solvent institutions. Prior to 1991, other federal regulatory agencies had successfully protected their turf, [FN40] leaving the FDIC only a modest role in regulating solvent state non-member banks and administering the deposit insurance system. [FN41] The 1991 "prompt corrective action" provisions, discussed in Part V, gave the FDIC increasing powers over all insured institutions as they approach insolvency, perhaps setting the stage for continued expansion of the agency's power in the future. [FN42]

B. The Current Superpowers of the Banking Agencies

One aim of this Article is to draw attention to the remarkable group of superpowers that have accumulated in the hands of the FDIC and RTC. This Section describes the current superpowers. Much of the rest of the Article examines why they exist, and whether they should exist in their present form.

Structurally, bank insolvency is special because the government agency that charters a bank holds a monopoly power to declare insolvency. [FN43] The structure also requires the FDIC and RTC to play many of the roles split in bankruptcy among the creditors, trustee, and bankruptcy judge.

The FDIC and RTC have assumed a remarkable list of superpowers once insolvency is declared. This Article selects two prominent types of superpowers for particular attention-cross-guarantees and D'Oench powers. They illustrate how systemic effects, insider abuse, and public choice can assist a normative assessment of special insolvency rules. The other special insolvency rules are described more briefly, giving the reader a sense of the pervasive way in which different rules apply before and after a bank's insolvency. Where helpful, the special rules are contrasted with bankruptcy law in order to highlight the unique nature of these agency superpowers.

1. Cross-Guarantees. Under the cross-guarantees provision of FIRREA, [FN44] the FDIC gained new and strong powers with respect to banks and thrifts affiliated with an insolvent institution. The provision requires that any loss incurred by the FDIC due to the failure of one institution is automatically assessed against affiliated banks and thrifts. Shareholders and non-bank affiliates of the failed bank are not required to pay, but their claims against the failed bank are subordinated to the FDIC's claim. These insiders will often not get paid even for secured claims against the failed bank. [FN45]

In bankruptcy, no special power exists for creditors to reach the assets of affiliates. To reach those assets, creditors must succeed on claims that are available outside of bankruptcy, such as fraudulent conveyance, equitable subordination, or piercing of the corporate veil. [FN46] Each of these claims can be laborious, and often impossible, to prove.

2. D'Oench Powers. Perhaps most surprising of all the agency superpowers are the D'Oench powers, which bar many claims and defenses against conservators and receivers that would have been valid against the bank itself. Three interrelated sources of law aid the agencies in increasing the value of the failed institution. First, the D'Oench case announced an equitable doctrine that prevents a person doing business with a bank from benefitting from any "secret agreement" undocumented in bank records and thus not discoverable by bank regulators. [FN47] Second, section 1823(e) of Title 12, a strict statutory version of the D'Oench doctrine, creates a specialized statute of frauds that defeats all claims or defenses against a bank except those based on contemporary approval by the bank's board of directors. The approval must also be continuously maintained in bank records. [FN48] Finally, the "federal holder in due course" doctrine has developed judicially to allow agencies to win in related cases even where D'Oench and section 1823(e) may not apply. [FN49]

The upshot of these powers has been agency victories over borrowers, even in quite extreme cases, such as where the bank unilaterally altered the terms of a note, [FN50] or forged the signature on a loan renewal and diverted the funds to a bank officer's use. [FN51] D'Oench cases most often involve a borrower's defense against FDIC or RTC enforcement of a loan. The agencies routinely win, usually on summary judgment, even where the claim or defense would have been valid under state or federal law if proved. To date, the agencies have won against a notable list of claims and defenses that includes fraud in the inducement; [FN52] various forms of misrepresentation; [FN53] various lender liability claims; [FN54] recklessness or negligence; [FN55] material alteration of a promissory note; [FN56] inaccurate recital of note terms; [FN57] unrecorded agreement; [FN58] improper notarization; [FN59] mechanic's lien; [FN60] failure of consideration; [FN61] accord and satisfaction; [FN62] novation; [FN63] unjust enrichment, waiver, and estoppel; [FN64] duress; [FN65] lack of mental capacity; [FN66] usury; [FN67] violation of federal securities laws; [FN68] and setoff rights of participating bank. [FN69]

3. Other Special Insolvency Rules. A substantial number of other special insolvency rules have developed in addition to the cross- guarantees and D'Oench powers. To understand this Article, it is not necessary to understand the intricacies of these special rules, many of which are still being contested in the courts. It is important, however, to grasp the general point that bank insolvency law overall is quite distinctive, and varies both from the law applying to banks pre-insolvency and from the insolvency law applying to non-bank corporations. [FN70] The reader may wish to review these special powers quickly, and refer back to this Section when the powers are discussed later in the Article.

The largest category of special bank insolvency rules singles out those with some insider connection to the failed bank and attempts to shift the costs of failure from the insurance fund to the insiders. In this category fall special rules concerning: repudiation of contracts and leases, fraudulent conveyance, asset freezes, priority among creditor claims, agency enforcement powers, and bankruptcy. Other special rules include: a unique definition of when a bank becomes insolvent, rules exempting the FDIC and RTC from state and local taxation, and special procedural rules.

A bankruptcy trustee or debtor, subject to the court's approval, has broad powers to repudiate executory contracts or unexpired leases. [FN71] This power prevents "sweetheart" contracts that allow insiders to get higher priority than other claimants. The banking agencies have essentially the same powers as the trustee, but with two material additions. First, the agencies appear to retain almost unlimited discretion about whether to repudiate, [FN72] in contrast to the bankruptcy court's supervision of a trustee's decision. Second, the agencies need not pay the full measure of contract damages that repudiation in bankruptcy requires; [FN73] instead, the receiver or conservator is excused from payment for lost profits on contracts and for future rent on leases. [FN74]

Three aspects of the fraudulent conveyance provision of the Crime Control Act of 1990 [FN75] illustrate the extraordinary powers granted to the FDIC and RTC by that Act. First, the FDIC and RTC can avoid fraudulent transfers made five years before or after appointment of the receiver or conservator. [FN76] Second, the agencies may avoid transfers not only of the bank itself, but of all "institution-affiliated parties," a broad term including any director, officer, employee, controlling stockholder, or agent for a bank. [FN77] Third, the rights of the agencies in the fraudulent transfer are superior to any rights of a bankruptcy trustee for the fraudulent transferee. The agencies will thus stand ahead of all other claimants, except federal agencies, on the estate of that transferee. [FN78]

The Crime Control Act of 1990 gave the FDIC and RTC the power to request a court to freeze the assets of any person, with the assets held by a trustee. [FN79] Recovery on the assets goes to the estate of the failed institution. The agencies may seek this prejudgment attachment under the injunction provisions of Rule 65 of the Federal Rules of Civil Procedure, but with one remarkable change: They are not required to make the usual showing that the injury is "irreparable and immediate." [FN80]

FIRREA granted the FDIC and RTC the power to discriminate among claims on the failed institution, in contrast to the traditional rule that all claims in the same class must receive the same percentage recovery. [FN81] The agencies are obligated to pay only the amount a claimant would have received if the agencies had liquidated the assets and liabilities of the institution, [FN82] which is often a steep discount from full recovery. [FN83] The agencies also retains discretion to select some claimants for additional payments, [FN84] thus enabling the agencies to treat insiders less well than other claimants.

Government agencies have wide-ranging enforcement powers both before and after insolvency, with no close analogues to creditor powers in bankruptcy. Notably, the FDIC and OTS can levy civil money penalties of up to $1 million per day for violations of banking rules. [FN85] And temporary cease-and-desist orders can freeze assets of institution-affiliated parties upon a mere prima facie showing that money penalties are appropriate. [FN86] Such enforcement actions have been allowed to go forward against a holding company despite the automatic stay provisions in bankruptcy. [FN87] Such enforcement proceedings can obviously enhance the ability of agencies to recover for losses arising from insolvent institutions. [FN88]

The Crime Control Act of 1990 also gave the agencies new powers that trump rules otherwise applying in bankruptcy. In order to ensure that bank insiders do not benefit from their acts, the Act declared certain banking- related claims no longer dischargeable in bankruptcy. [FN89] In addition, the Act created new rules that apply to any entity, such as a bank holding company, that commits to maintain the capital of an insured bank. If the holding company later becomes bankrupt, the FDIC is given high priority to the amount of the commitment. [FN90] These rules became more important in 1991 with passage of FDICIA, which routinely requires such commitments by holding companies to maintain the capital of undercapitalized banks. [FN91]

In 1991, FDICIA established a regime for prompt corrective action for undercapitalized banks. A series of increasingly stringent rules now apply to banks as their capital levels fall below statutorily required levels towards insolvency. Most important for purposes of this Article, the Act introduces a new trigger for bank insolvency. Instead of waiting for banks to exhaust their capital, the Act requires that "severely undercapitalized" banks (those whose capital is less than two percent of assets) be closed within ninety days of becoming severely undercapitalized. [FN92] The bank is thus subject to the full range of agency receivership powers even before the usual definitions of insolvency are met. One solid rationale for this rule is that, due to accounting imperfections, banks reporting two-percent capital are actually likely to be insolvent. [FN93]

The FDIC and RTC acting as receiver are exempt from all taxation imposed by any state or local taxing authority, except for certain real property taxes. [FN94] These agency advantages both reduce the costs to the insurance fund of a receivership, [FN95] and, echoing McCulloch v. Maryland, [FN96] prevent the FDIC and RTC from being subjected to state and local taxation.

The FDIC and RTC enjoy a number of special procedural rules that a bank would not have available pre-insolvency. The agencies have a broad right to remove to federal court, [FN97] and can remove a suit from state court without bond. [FN98] They can stay litigation against the bank for ninety days in the case of receivership and forty-five days for conservatorship. [FN99] These last powers are similar, although far from identical, to the automatic stay in bankruptcy, [FN100] and the use of a federal bankruptcy court to replace state proceedings by creditors. Additional advantages to the agencies acting as receiver or conservator are an extended statute of limitations, [FN101] and broad investigative powers, such as the ability to issue a subpoena or subpoena duces tecum. [FN102]

C. Summary

Since well before the creation of deposit insurance, receivers for failed banks and thrifts have enjoyed powers that the institutions would not have had pre-insolvency. In recent years, the number of these "superpowers" has grown substantially. The remainder of this Article seeks to explain the development of these special rules, and to assess their desirability.

II. SYSTEMIC EFFECTS OF BANK FAILURE AND SPECIAL BANK INSOLVENCY RULES

The unusual problems of bank failure-contagious bank runs and the blocking of depositors' immediate access to their funds-have led to special rules to prevent systemic losses from bank failures. The creation of the FDIC and deposit insurance has provided new rationales for special bank insolvency rules, such as saving the insurance fund and deferring to FDIC expertise. This Part examines the persuasiveness of these rationales for special rules, concluding that the current rules are substantially more favorable to the FDIC and RTC than is justified by concern for systemic losses.

A. Special Bank Insolvency Rules Before Deposit Insurance

An initial question is why special insolvency rules originally developed. In the period before deposit insurance, a bank failure posed two sorts of threats to the local or national economy. First, depositors would lose access to their transaction accounts-their checks would not clear, and they would in general lose the key benefit of "immediacy," or immediate access to their money. [FN103] Thus, depositors would prefer a deposit contract that protected immediacy. As discussed below, special bank insolvency rules have done precisely that.

The second threat posed by the failure of a bank was that failure would spread to other banks in a contagious bank run. Observers in the nineteenth and early twentieth centuries had no doubt about the bad effects of bank runs. Consider this vivid description by a Kansas court:

[In case of insolvency,] the mischief takes a wide range. Those who have been accommodated with loans must pay, whatever their readiness or ability to do so. Further advances cannot be obtained. Other banks must call in their loans and refuse to extend credit in order to fortify themselves against the uneasiness and even terror of their own depositors. Confidence is destroyed. Enterprises are stopped. Business is brought to a standstill. Securities are enforced. Property is sacrificed, and disaster spreads from locality to locality. All these incidents of the banking business are matters of common knowledge and experience. [FN104]

This description shows how bank runs were thought to be tightly linked with the three main reasons that banks historically have been considered "special" and thus deserving of a special regulatory regime: "1. Banks offer transaction accounts. 2. Banks are the backup source of liquidity for all other institutions. 3. Banks are the transmission belt for monetary policy." [FN105] Transaction accounts have already been discussed in terms of depositors' loss of immediacy; the Kansas court's description suggests how bank runs would squeeze businesses that had transaction accounts, forcing them to repay loans "whatever their readiness or ability to do so." To fortify their own balance sheets, banks would cease providing backup liquidity for (that is, stop lending to) other institutions. Finally, banks' reluctance to extend credit would contract the money supply, as effects of the run spread "from locality to locality."

The threat of depositors' loss of immediate access to funds, and the fear of contagious bank runs, led early on to a special bank insolvency regime. Three sorts of insolvency rules were developed which reduced the likelihood of bank failures by lowering the incentive of depositors to begin runs. [FN106]

First, depositors could take comfort in substantial extra capital and liquidity that was likely available to a bank in distress or insolvency. National banks and many state banks operated under the double liability rule. [FN107] Substantial assistance was also often available from a bank's clearinghouse, [FN108] or, after 1913, from the Federal Reserve. The likelihood of fresh funding, in turn, made it less likely that a creditor would feel insecure and be tempted to start or participate in a run.

A second way to protect immediacy and stem contagion was to assure that bank insolvencies were handled as quickly as possible. A short interruption in access to funds had lower costs to depositors and induced less panic than the likelihood of the long, often multi-year, receiverships common in non-bank settings. The use of statutory receivers helped to speed up liquidations by allowing administrative determinations of many claims without time-consuming recourse to the courts. Statutory receivers did not have their salaries and expenses paid out of the estate, so they did not have the usual incentives to stretch out the receivership in order to increase their remuneration. In addition, statutory receivers early made a practice of partial distributions to depositors so that bank creditors could expect to get much of their funds quickly. [FN109]

A third way to protect depositors' access and stem contagion was to give the government a monopoly on declaring a bank insolvent. In the era before deposit insurance, a run might have started on the mere rumor that a creditor was seeking to have a bank declared insolvent. A creditor therefore could have strategically threatened to seek such a declaration. The bank then would have faced the unpalatable choice of either giving uneconomic concessions to the creditor or facing a run. The existence of this credible threat by a bank creditor may explain passage of "bank libel" laws, which made it a crime to threaten to spread rumors about a bank or otherwise cause a bank run. [FN110] In states where the government did not have a monopoly on the closure decision, a great deal of litigation arose over whether banks were insolvent, creating another way for confidence to be shaken. [FN111]

The avalanche of bank failures during the Great Depression [FN112] led to widespread political support for drastic changes in the banking system. The FDIC was created in 1933, and the FSLIC in 1934, based on the political belief that federal deposit insurance was the best way to raise confidence in the banking system and restore liquidity to the economy. [FN113] The existence of deposit insurance created new reasons for favoring special rules in bank insolvency.

B. Assessing Immediacy and Bank Runs as the Bases for a Special Insolvency Regime

Although concerns about immediacy and bank runs explain the early special bank insolvency rules, it is not clear that they remain convincing rationales for special rules today. The pre-1933 special rules for bank insolvency evolved against a background of the general rule for corporate insolvencies, under which the relative substantive rights of creditors are rarely altered by the fact of insolvency. [FN114] Modern bankruptcy scholarship has emphasized how this general rule solves a collective action problem facing creditors. In times of business difficulty, creditors seek to assure their interest in the debtor's assets. Creditors having a preferred status in bankruptcy will seek to throw the debtor into bankruptcy. Creditors lacking the preferred status will seek to grab assets before bankruptcy is declared. As a result, the collective group of creditors will be less well off-creditors will expend resources trying to grab the debtor's assets, and the going concern value of the debtor may well be destroyed when it is placed in bankruptcy. Under the "creditors' bargain" model advanced by Thomas Jackson and Douglas Baird, [FN115] the creditors as a group should prefer a bankruptcy regime where the same substantive rules apply in and out of bankruptcy. [FN116] Having the same rules will reduce the incentive for any creditor to begin the grab for assets.

Such a race to grab assets is very familiar in the banking context-it is known as a "bank run." [FN117] Creditor runs are more likely to occur in banks than in other corporations due precisely to banks' reliance on transaction accounts, from which depositors have the right to withdraw their money immediately, at full par value. These depositors have shown a high demand for immediacy, [FN118] and will lose the benefit of their bargain if they cannot gain access to their funds. Furthermore, because banks keep only a fraction of deposits on hand, banks cannot pay depositors' demands if a large fraction of depositors demand payment at once. Banks thus face a more acute form of the collective action problem than is handled by the bankruptcy system. An individual depositor, like an individual corporate creditor, faces the risk that the debtor will not have enough assets. Depositors or creditors thus have an incentive to grab their own assets while it is possible to get full recovery, triggering a rush for assets that will result in insolvency, a smaller pool of assets available for creditors as a group, and reduced recoveries for those who did not grab quickly enough. The collective action problem is more acute for banks because most depositors have the right to demand immediate payment, whereas many corporate creditors can demand payment only over a longer period.

Special insolvency rules are one way to reduce the likelihood of such bank runs. A consistent theme of bank insolvency rules has been to provide assurances to depositors that additional funds will be available to the bank in times of stress. That assistance might come from double liability from shareholders, or support from the bank's clearinghouse, or a guarantee by a credible insurer such as the FDIC. In each instance, the credible promise of extra funding reassures depositors, confirms their immediate access to transaction accounts, and lessens their incentive to begin a destructive bank run.

A number of observations, however, may lead one to doubt that the collective action problem justifies today's special insolvency rules only for banks and not for other corporations. The problem of bank runs explains the need to assure depositors of additional funding, but does not explain the many other special bank insolvency rules described in Part I. Over time, the distinctions between banks and other corporations have eroded significantly, with non-banks today often reliant on short-term financing similar to deposits. [FN119] In addition, recent scholarship has cast doubt on the traditional understanding of bank runs exemplified by the Kansas court's description. [FN120] The crucial empirical issue related to bank runs is how likely it is that a run on a particular bank will become a generalized run on many banks. If this likelihood is low, the macroeconomic effects of an individual bank run will be trivial, as depositors will simply transfer their funds to a solvent bank, which will continue to finance commercial activities. If, however, this likelihood is high, the macroeconomic effects of an individual bank run could be substantial-lower business activity and reduced money supply.

George Kaufman has recently presented evidence that the macroeconomic effects of bank runs have been exaggerated, and that nationwide bank contagion probably occurred only in 1893 and during the Great Depression. [FN121] If the risk of contagion is indeed low-a conclusion not shared by regulators and others [FN122]-then the historical reasons for special bank insolvency rules become much less important.

This Article does not attempt to answer the debate about the true risks of bank runs. The greater these risks, the greater the benefits of special insolvency rules, such as deposit insurance, which reduce the costs of bank runs. These benefits were likely substantial in the period after 1933, when creation of deposit insurance drastically reduced the number of bank failures by reducing the likelihood of runs. [FN123] Over time, however, these benefits of deposit insurance have carried with them an increasing cost-the cost to the taxpayers of paying for higher numbers of failures. One major source of these failures is "moral hazard," or the incentive for banks covered by insurance to act differently because of that insurance. [FN124] Institutions have taken advantage of freely available and underpriced deposit insurance, and have taken on excessive risk secure in the knowledge that they would profit from the upside and the insurance fund would pay for the downside. Deposit insurance thus has reduced costs of depositor runs, but has increased costs of excessive risk taken on by institutions.

C. Implications of Deposit Insurance for Bank Insolvency Rules

Resolving the tension between reducing runs and reducing deposit insurance payouts is beyond the scope of this Article. At present, deposit insurance remains in place. Trying to reduce the costs of payout from the insurance system has become a major reason for broad regulatory powers both before and after insolvency. The next question is which powers are explained or justified by the presence of deposit insurance.

1. "Saving the Fund" as a Source of Special Rules. Perhaps the simplest argument for special powers for the FDIC is the "save the fund" argument: The FDIC insurance fund pays the cost of failures covered by deposit insurance; [FN125] rather than having taxpayers suffer the losses, those connected to the bank failures ought to pay. [FN126] Under this argument, the rules should favor protection of the deposit insurance fund, and perhaps on close legal issues some benefit of the doubt should be given to the FDIC. I conclude that although it deserves some weight, the argument by itself does not justify anything like the long list of current superpowers.

a. Reasons supporting the "save the fund" argument. The structure of deposit insurance, equitable considerations, and special difficulties facing government collection efforts all may support the "save the fund" argument.

The structure of deposit insurance puts greater burdens on government than on a similar private insurer, and thus may justify certain benefits for government in being reimbursed. Because termination of deposit insurance is considered the equivalent of a death penalty for banks, that penalty is rarely used. [FN127] The FDIC insurance fund has thus become a sort of "assigned risk" pool, with the government required as sole insurer to take on all the bad risks.

Not only has the FDIC had to insure the bad risks with the good, but it has been saddled by statute with a product that no private insurer would tolerate. The deposit insurance program is in the form of a limitless guarantee-insolvent institutions place their full negative net worth into the insurer's hands. By contrast, liability in private insurance contracts is carefully capped. The rules governing rights of private creditors thus may not be appropriate for determining the rights of the government when it seeks reimbursement for its extraordinary guarantee.

The unlimited federal guarantee creates moral hazard in a more acute form than would be created in a private insurance market. Banks can take on a limitless amount of insured deposits. Moreover, as discussed below in Part III, private-sector monitoring is less effective for banks than for other corporations, in part due to the presence of government monitoring. Finally, the usual private-sector protections of the insurer, such as deductibles or co- payments by the insured, do not apply in FDIC insurance. The FDIC is now beginning to set premiums based on the riskiness of the bank, [FN128] which may be a helpful first step toward mimicking private-sector mechanisms that limit moral hazard. But as long as government effectively lacks the private sector's ability to terminate a bad risk's insurance, the FDIC will be worse off than a private insurer. Banks will continue to take on risk to the extent that they can escape detection by heavily burdened regulators.

The unlimited guarantee, combined with acute moral hazard, suggests that the deposit insurance fund is more vulnerable to large losses than analogous private insurers. Yet the disadvantages of the current system, such as the limitless nature of the guarantee and the absence of deductibles and co- payments, are precisely the features that best assure depositor confidence in the banking system. The government thus can make a sort of "sweet with the bitter" argument for special protections when the insurance fund pays out on failed banks.

Another reason the government may deserve special rules in insolvency is that equity may favor it over the third parties with whom it is competing for assets. Imagine, for instance, that the other creditors are mostly insiders who have received fraudulent transfers from the bank before insolvency. Such a situation might justify special rules for fraudulent conveyance in bank insolvency, favoring the insurance fund. More generally, as discussed in Part III, special rules may be appropriate to the extent there is evidence that the government is equitably in a stronger position than insiders who may have acted abusively.

There may also be characteristic difficulties that face government more than private creditors. Special rules in insolvency may balance out these government disadvantages. One disadvantage is that government agencies may, on average, be less successful at collecting assets than private creditors. Private creditors often have direct financial stakes in realizing on amounts due, while government employees may have a different set of incentives less clearly tied to recovering money. [FN129] There is also less reason to think that government invests efficiently in collecting amounts due. Whereas a private creditor should invest in collection until the marginal cost from collection efforts equals the marginal return, the level of government collection efforts may be driven by other considerations, such as budgetary rules that limit the number of employees, contracting rules that make it costly for the agencies to adjust the level of collection effort, and political concern that the government appear aggressive in collection efforts.

An additional disadvantage is that the government will often be the last of the creditors to be preferred when the debtor has discretion about making partial payments. A debtor facing bankruptcy, including a bank manager, has an incentive to favor creditors who can be of benefit post-insolvency. There is no similar incentive to favor the government, which is unlikely to show favoritism post-insolvency. [FN130] This government disadvantage may justify special rules to compensate for situations where other creditors are favored pre-insolvency. More important, this disadvantage is a good description of how deposit insurance works generally-banks and their managers may choose to benefit anyone else before worrying about incremental losses to the deposit insurance fund. The fund may thus deserve special protection in bank insolvencies.

b. Reasons for rejecting the "save the fund" argument. One strong reason for restricting the "save the fund" argument is that it is too open- ended. If the actual goal is to maximize recovery by the FDIC on bank assets, then quite extraordinary measures might be considered. For instance, the FDIC might be able to defeat all claims against the bank except those of insured depositors, with the fund recovering all remaining money. Or, all defenses against collection by the FDIC might be disallowed. (The D'Oench powers move far in that direction.) Bank managers or employees might be made to disgorge some or all of their compensation in the year before insolvency, and so on. But it cannot be the case that the mere fact that the government pays means it can do whatever it wants. Additional considerations must be brought to bear on the decision of how far the "save the fund" argument should go.

A second limit on the argument is that it makes little sense when analyzed as tax policy. The costs of the deposit insurance system are spread over a wide base, with the assessments directly made on deposits, and backup funding coming from taxpayers. All depositors and taxpayers receive the macroeconomic benefits of a smoothly functioning banking system. By contrast, special rules that favor the government come at the expense of an unknown group of private parties. The rules thus operate as a sort of random tax on parties unlucky enough to be caught by the special rule. [FN131] Absent some showing that the losing parties deserve to be singled out for such taxation, the better tax policy would seem to be to spread the cost over the full range of beneficiaries of the deposit insurance system. [FN132]

The equitable position of the government is not nearly as strong as suggested above. Government agencies have pervasive control over banks, and their intervention in bank management often intensifies as the bank nears insolvency. [FN133] The general rule is that a creditor having dominion and control over the debtor is in a weaker position than other creditors. [FN134] The government, because it is in a position to control the bank, equitably should come behind other creditors who do not have the benefits of similar control. The equitable position of the government has thus weakened as bank regulators have gained sweeping new powers in recent years.

A related challenge to the government's equitable position comes from the "creditors' bargain" model. As discussed above, Jackson and Baird are critical of special distributional rules in bankruptcy because of the possibility that a creditor might force bankruptcy to take advantage of such rules. [FN135] Because of its monopoly on bank closure, only the government would similarly be in a position to take advantage of rules in bank insolvency. To the extent that special rules give the government an incentive to alter its decision on whether to close a bank, other creditors are disadvantaged, and thus in an equitably superior position to the government. This risk of the government advantaging itself has become greater under the new early closure rules. The government will now close banks sooner to ensure higher recovery on insured deposits-to the detriment of shareholders and others who lose value when a still-solvent bank is declared insolvent. Part V explains how this loss to shareholders and others may, in the long run, actually result in more failures and greater eventual costs to the government.

c. Assessing the "save the fund" argument. On close inspection, it is far from clear that there should be a general presumption in favor of the government and against other parties affected by a bank's insolvency. Although the government has special difficulties in collecting on receivership assets, limits must be set on the powers that the government can claim in order to save the fund. Although the government has plausible equitable claims, so do many third parties. Part III examines further how to analyze these equitable claims, which are linked to the suspicion that bank failures are caused by insider abuse. Part IV discusses a sharper reason for being suspicious of the "save the fund" argument. Both the FDIC and Congress have strong political incentives to err on the side of saving the fund, and placing costs on third parties. These political incentives, rather than sound public policy, may be the strongest explanation for many of the recent special insolvency rules favoring the fund.

2. FDIC Expertise as a Source of Special Rules. The presence of deposit insurance may justify special insolvency powers for the FDIC, in addition to whatever powers it deserves in order to save the insurance fund. [FN136] The FDIC is systematically different from the ordinary creditor in a bankruptcy case. In every bank insolvency the agency becomes the dominant creditor when it takes over the claims of insured depositors, typically having over ninety percent of total claims against the receivership. [FN137] In order to save the fund, therefore, the agency has good incentive to get substantial recovery on the bank's assets. The agency also is in the unusual position of being a pervasive, repeat player-every modern bank insolvency has the FDIC as receiver.

Three sorts of reasons support special FDIC insolvency powers in light of its unique role as dominant creditor in each of the many bank insolvencies. First, and most generally, the FDIC can make a plausible claim to expertise. With high volume comes a variety of economies of scale-procedures are instituted, manuals are developed, and specialized skills are developed. In short, the FDIC can claim the advantages of a bureaucracy devoted to a repetitive task. Second, the FDIC can seek consistency among cases. Its central role in repeated insolvencies means that it can aspire to the goal of distributive justice, of treating like cases alike while making justified distinctions. Third, the FDIC can seek a logical development of the law consistent with public policy. Because the agency is involved in so many cases, it can choose appropriate instances to test principles and develop precedent. The FDIC's cases are heard in federal courts, where there is opportunity to develop law oriented toward a broad range of legal concerns; by contrast, bankruptcy courts are likely more oriented toward the adjustment of the rights of a particular set of debtor and creditors.

These possible advantages may justify the FDIC's unusual combination of the powers of a bankruptcy judge, creditor, and trustee. The advantages of bureaucracy and expertise may justify the FDIC's power to adjudicate claims against the receivership. [FN138] As creditor, the FDIC insurance fund appropriately receives high priority on recovery of the bank's assets. And as receiver, the FDIC may deserve its range of powers, such as the ability to repudiate contracts and to seek collection in federal court.

Although the FDIC's justifications are forceful, there is reason to be cautious about approving the full range of current powers. The central concern is how well the FDIC will employ its enormous discretion. The range of its discretion is easily illustrated: what means to use in resolving a failed bank; [FN139] whether to issue a policy statement that certain categories of contracts will not be repudiated; [FN140] whether to grant an exception to the cross-guarantees power; what sorts of D'Oench powers to apply; and, more generally, when to seek penalties against individuals and institutions for alleged violations of banking law.

The unique role of the FDIC as dominant creditor and repeat player offers a possible explanation and justification for broad discretionary powers for the agency in bank insolvencies. Judging these powers, however, requires a difficult empirical estimate. If the FDIC generally uses its discretion well, then bank insolvency will be marked by expert administration, distributive justice, and orderly development of desirable law. If that discretion is abused, however, then bank insolvency can descend into arbitrary administration, favoritism or vindictiveness, and a cynical and selective prosecution of cases in violation of public policy. Agency discretion also creates uncertainty for third parties, which, as discussed further in Part V, can have a chilling effect on those doing business with banks. As our experience with modern bank insolvency grows, Congress, the courts, and the agencies should look for appropriate opportunities to confine FDIC discretion, such as through more rulemakings and agency statements of policy.

D. Summary

The existence of special bank insolvency rules was founded on the perceived problem of bank runs and depositors' need to have immediate access to transaction accounts. Whether or not the costs of bank runs are as great as commonly believed, the current deposit insurance system was created as a means of stopping runs. It has largely succeeded in that goal, but has subjected taxpayers in recent years to the enormous costs of the bailout, due in part to the extra risks taken by institutions covered by insurance. If something like the current system is maintained, then its very existence creates arguments for special bank insolvency rules based on "saving the fund" and agency expertise. Although these arguments have some force, they are overbroad, and do not explain just when the fund should be saved or the agency given discretion. Moreover, the arguments for special bank insolvency rules are subject to strong counter-arguments, considered further below.

III. THE PREVALENCE OF INSIDER ABUSE AND SPECIAL BANK INSOLVENCY RULES

Many of the special bank insolvency rules can be explained, and perhaps justified, by the peculiar likelihood that a bank failure is caused by the misdeeds of insiders. Imagine at the extreme that every bank failure is caused by "insider abuse." "Insider abuse," as the term is used here, refers to two distinct sorts of behavior by insiders, who are management and any shareholders who play an intimate role in directing the corporation. The first sort is fraud or self-dealing-insiders taking advantage of their position to benefit themselves in unpermitted ways. In corporate law terms, this behavior is likely to violate the duty of loyalty to the corporation. [FN141] The second sort of behavior is severe mismanagement, which means behavior reckless or stupid enough to be second-guessed by courts, even under the generous deference of the business judgment rule. [FN142]

By contrast, imagine that every failure of a non-bank is the result of stiff competition in a competitive market. In these failures, there is no severe mismanagement or violation of the duty of loyalty. Instead, management's decision was reasonable ex ante. At the time of the decision, there was a probability distribution of results, and the expected value of this distribution exceeded the expected value of the alternatives that were open to management. Sadly enough, the corporation experienced the negative tail of the distribution, and therefore failed. More generally, failure might result from any management decision that was good enough to survive scrutiny under the business judgment rule.

This Part argues, as a descriptive matter, that current insolvency law has been strongly influenced by the "model of insider abuse" for banks and thrifts (and sometimes other financial institutions) and by the "model of stiff competition" for other corporations. In particular, many of the strict rules enacted in 1989, 1990, and 1991 conform to the "model of insider abuse."

Although this model helps explain the recent passage of strict rules, it has become dated. The available evidence suggests that banks and thrifts today face much stiffer competition than in years past, so that current failures are more likely to be due to competition rather than abuse. Nonetheless, from a theoretical perspective, there continue to be factors which make insider abuse more likely in banks than in industrial corporations, including the relative ease of doing fraudulent transactions and the relative weakness of monitoring of banks by the private sector. To the extent that it is demonstrated that certain forms of insider abuse are especially likely in the banking context, some special rules for bank insolvency may be justified. But the volume of recent legislation raises the strong possibility that more rules have been passed than can be so justified.

A. The Historical Basis for the "Model of Insider Abuse"

The history of banks and thrifts since the adoption of deposit insurance suggests that the "model of insider abuse" described bank and thrift failures fairly well from the creation of deposit insurance in 1933 until about 1980. Numerous rules limited competition and the risk of insolvency. [FN143] Glass- Steagall [FN144] and other restrictions forced institutions to specialize by product-banks, thrifts, securities firms, and insurance companies were kept off of each other's turf. Banks faced pervasive limits on geographic expansion. For instance, branching within a state, or even within a county, was often prohibited. [FN145] Entry to banking was carefully controlled by regulators who denied many charter applications. Finally, rate regulation strictly controlled the interest banks and thrifts could offer on savings accounts, and forbade altogether interest on checking accounts. [FN146] Local banks thus enjoyed substantial monopoly power, free from the worry of price competition or new entry.

With the protection afforded by these rules, it often took fraud or severe mismanagement for a bank to fail-and thus the "model of insider abuse" appears to have been a good fit until about 1980. In 1976, when much of the monopoly system of regulation was still in place, Robert Clark reported on the relative frequency of insider abuse in banks and non-banks. [FN147] The FDIC reported to Clark that for the 80 insured bank failures between 1960 and 1975, the basic causes of failure were insider loans in 42 cases (52.5%) and insider misappropriation in 24 cases (30%). [FN148] The remaining failures were attributed to managerial weaknesses. None of the failures were attributed primarily to business downturns or other market-based causes. [FN149] By contrast, Clark reported that "only a negligible proportion of the business failures of ordinary industrial corporations is due to fraudulent and self- dealing conduct." [FN150]

The hundreds of thrift failures of the 1980s have been portrayed popularly, and in Congress, as due to a new wave of insider abuse. On this account, when the thrift industry blamed overregulation for the losses of the late 1970s and early 1980s, Congress responded by granting thrifts broad new powers and raising the insured amount from $40,000 to $100,000 per account. [FN151] Expanded powers, combined with a simultaneous sharp reduction in thrift supervision, opened opportunities for abuse. The moral hazard created by deposit insurance, and newly fed by brokered deposits, meant that thrift operators could bet enormous sums on a "heads I win, tails the government pays" strategy. Adverse selection also occurred, as sharp operators entered the business to take advantage of the chance to gamble large amounts of other people's money. For example, Charles Keating was allowed to buy a thrift even though he had a cease-and-desist order outstanding from the SEC for previous bank-related abuses. [FN152]

The flamboyant scandals surrounding thrift owners such as Keating and David Paul [FN153] gave Congress a plausible basis for acting on the perception that thrift failures were due to insider abuse. FIRREA, the Crime Control Act of 1990, and FDICIA had numerous strict provisions that can be understood as attempts to attack "S&L kingpins." [FN154] These laws, such as innovative ways to prevent institution-affiliated parties from transferring funds away from regulators, make a good deal of sense if bank and thrift failures indeed systematically arise from insider abuse.

Ironically, however, the strict laws of recent vintage have been passed just as the "model of stiff competition" has become a much stronger general explanation for bank and thrift failures. In the 1980s, the old monopoly system broke down. Insurance companies, securities firms, thrifts, banks, and "non-bank banks" invaded each other's turf. Branching rules were relaxed. New entry increased as regulators became more likely to approve charter and merger applications. [FN155] And rate regulation was abandoned in response to high interest rates and the development of money market mutual funds as an attractive alternative to bank and thrift deposits. In short, banks and thrifts since the early 1980s have faced sharply increased competition, as well as the strain of high and volatile interest rates and deep regional recessions.

As monopoly profits eroded, banks and thrifts lost their previously enormous margin for error. In their newly competitive markets, a few bad management decisions can be enough to result in failure. These more challenging external conditions have, like a low tide, left more and more institutions stranded in agency hands. The recent history suggests that many, and perhaps most, bank and thrift failures have been attributable to stiff competition. [FN156] In the future, the strict new criminal and civil enforcement rules should further deter insider abuse, while making failure for business reasons relatively more likely. [FN157] It has become anachronistic to assume that failures are due to insider abuse.

B. The Continuing Tendency of Insider Abuse in Bank Failures

Although bank failures are today less likely to be due to insider abuse, such abuse is still more likely to occur in banks than industrial corporations. A maxim has it that the best way to rob a bank is to own one. Fraud is easy to conduct; information for controlling fraud is difficult to come by; and usual means of monitoring are weaker than for industrial companies.

1. Fraud and the Nature of Banking Transactions. Banking involves transactions that are (1) numerous, (2) in highly liquid form, (3) easily forgeable, and (4) involve large amounts of money which (5) often cross jurisdictional boundaries. Each of these factors make it easier for insiders to steal. First, a large volume of transactions, such as the mortgages and commercial loans that flood through a bank office, is harder to monitor than a small number. It is easier to get away with an occasional fraudulent deal set amidst a large number of legitimate ones. A large number of transactions can also make detection difficult, such as when assets are shuttled through numerous dummy corporations in a series of complex transactions. Second, liquidity is of great help to a thief-it is far easier to fence stolen dollars than a stolen machine press. [FN158] Third, financial transactions, such as loans, are easily forged or otherwise included in fraudulent schemes-it is easier to find a taker for fake loan documents than for fake machine presses. Fourth, significant fraud is better done in large transactions. An illicit $10,000 is more easily hidden in a huge real estate loan than in a used car loan. Fifth, crossing between jurisdictions often means that no one can adequately perceive the true recipient of the funds. [FN159]

These five factors can have a synergistic effect. The 1991 failure of the Bank of Credit and Commerce International (BCCI) is a type case of a huge insider fraud that escaped detection for years. [FN160] The "bank within a bank" manufactured a great number of transactions, involving liquid assets, that were both readily forgeable and for large sums of money, and shuttled among separate corporations in as many countries as possible. BCCI is an extreme case, but it remains plausible that insider fraud on a large scale is much more likely to occur in banks and similar financial intermediaries than in industrial corporations.

Although there has been little attention in the theoretical literature to explaining the special propensity for insider abuse in financial institutions, the legal system has long recognized the problem. In a variety of contexts, courts have placed a heightened duty on the officers and directors of financial institutions, [FN161] and have based this duty on the increased likelihood of fraud. The likelihood of insider abuse helps to explain the specialized regulatory regimes that have developed for all major categories of financial institutions, including pensions, investment companies, insurance companies, and securities firms. [FN162]

2. Information Inadequacy and the Likelihood of Insider Abuse. Problems of information inadequacy, which trouble stakeholders in corporations generally, are especially acute in banking. Following Charles Goodhart's analysis, the crucial problem is how difficult and costly it is for consumers or monitors of financial services to distinguish between banks with more, and banks with less, risky strategies. [FN163] The usual market signals are inherently ambiguous. A high rate of return may be offered by a bank because of greater efficiency. But it may also be offered because the bank is taking greater risks or practicing insider abuse. This ambiguity is made worse by the fact that a bank's rate of profitability is not a good predictor of future likelihood of failure-high-risk strategies often produce high returns in an early period before leading to spectacular failure later. [FN164]

The nature of commercial lending also makes it unusually hard to assess the likelihood of insider abuse and future solvency. The securities markets in recent decades have taken over a large fraction of the short-term financing traditionally funded by commercial banks. [FN165] The remaining commercial loans are predominantly to "middle-market" or small companies, or for other projects that are not readily financed by securities. [FN166] Middle-market loans are often complex, nonstandard contracts, embedded in a broader business relationship and difficult to price accurately. A variety of technical difficulties have thus far frustrated all attempts by regulators to compel banks to mark their loan portfolios to market. [FN167] Without mark-to- market, the value of a bank's loan portfolio is difficult for outsiders to evaluate, creating uncertainty about the ability of the bank to remain solvent in the face of market risks. More complete disclosure by banks of their loan portfolio would also compromise the confidentiality of information provided to the bank, one of the traditional advantages of borrowing from a bank. [FN168] Robert Clark has noted an additional source of information inadequacy: the relatively great ability of banks and similar financial institutions to shift their corporate strategy without notice to shareholders or bondholders. [FN169]

3. The Relative Weakness of Bank Monitoring. The increased likelihood of fraud and problems arising from information inadequacy would exist even in a deregulated banking system. These possibilities of insider abuse are compounded by the weakness in the usual types of monitoring compared with public, industrial corporations. Government monitoring is far more detailed with respect to banks, but still has characteristic weaknesses.

a. Depositor monitoring. The current deposit insurance system makes significant monitoring by depositors highly unlikely. Insured depositors have little reason to invest in monitoring the solvency of banks. After all, they are fully insured, and the FDIC has made a strict policy of making insured deposits available immediately in the event of bank failures. [FN170] Uninsured depositors have had little more reason to invest significant resources in monitoring. Most large bank failures have been handled through a purchase and assumption transaction, in which insured deposits are simply transferred to the acquiring bank. [FN171] In addition, the bulk of uninsured deposits are held in banks considered "too big to fail." Thus uninsured depositors are fairly confident that they will be protected by the FDIC's resolution of the failure; they are even more confident that they will be able to spot a potential large failure in time to move uninsured deposits to another bank. [FN172]

Much has been written on the role of depositor monitoring were deposit insurance to be abolished or substantially modified. [FN173] Advocates of market discipline have suggested that private institutions would develop to constrain bank risk-taking. If so, there would be less reason to suspect that bank failures arose from insider abuse. Opponents of market discipline have emphasized the cost of depositor monitoring, the increased risk of bank runs, and the disproportionate costs to less sophisticated depositors. In this Article, I do not choose sides in the debate over how to reform deposit insurance; thus far, major reform has been stymied. I note, however, that shifting large losses to depositors would reduce the stake of the insurance funds in bank failure, and consequently weaken "save the fund" rationales for special rules in bank insolvency.

b. Stockholder monitoring. The stock market likely plays a less effective role in monitoring management in banks than in industrial corporations. Outside bank shareholders suffer from the same information inadequacies as depositors. Bank shareholders also are likely disadvantaged by the special regime of bank securities regulation. The bank regulatory agencies, rather than the Securities and Exchange Commission, oversee bank securities issues, and there is evidence of less strict enforcement and less disclosure of potentially useful information. [FN174]

Nor has the market for corporate control operated as effectively with respect to banks. Hostile takeovers have been rare among banks and thrifts, with the Bank of New York's takeover of Irving Trust in 1989 the one notable exception. [FN175] Such takeovers can serve as an incentive for investors to discover badly managed banks; and the threat of takeover acts as a goad for managers to avoid being uncovered as bad managers. Without the threat of hostile takeover, bank management has historically been more free to engage in insider abuse, secure in the knowledge that bad management or fraud would not trigger an attempt by investors to select new management. [FN176]

c. Monitoring by secured creditors. An important body of theoretical literature has emphasized the leading role that secured creditors (often banks) can play in monitoring behavior by secured debtors. [FN177] Banks are free from this sort of monitoring. Their dominant source of secured credit is in the form of sophisticated securities transactions that take place in large, impersonal markets, rather than in relationships in which creditors monitor debtors.

d. Bondholder monitoring. Many corporations face significant monitoring from the owners of their debt, and bond covenants are an important way to control corporate risk-taking. [FN178] Bond covenants are not, however, an important way to monitor and control bank risk. Despite proposals to require banks to issue bonds, [FN179] banks currently issue only a small amount of them, in part because most bonds do not count toward a bank's required capital. [FN180] Unless market conditions or regulatory requirements change drastically, bondholders will continue to be far less important in the monitoring of behavior in banks than in industrial corporations.

e. Government monitoring. In light of the systematic weaknesses of private monitoring, government agencies have been assigned the major role in preventing insider abuse in banks and thrifts. (The existence of government monitoring may also have prevented effective private monitoring from developing.) Today, there is reason to believe that government monitoring is a better deterrent against major insider abuse than it was previously.

Over time, publicized incidents of abuse have led to agency requests for additional enforcement powers, which have eventually been granted. Examples of this pattern include the Financial Institutions Supervisory Act of 1966, [FN181] the Garn-St. Germain Depository Institutions Act of 1982, [FN182] FIRREA in 1989, [FN183] and the Crime Control Act of 1990. [FN184] The arsenal possessed by regulators today is truly formidable. For instance, fines of $1 million per day and the agencies' ability to freeze the assets of alleged wrongdoers should prove an effective deterrent to insider abuse. These new powers are accompanied by much stricter agency supervision than financial institutions, especially thrifts, received in the early 1980s. Indeed, as discussed below in Part V, the new agency powers may have become so strong as to chill legitimate banking activity. An important risk today is that over- strict insolvency rules will reduce the capital flowing into banking, thereby shrinking the size of the banking industry, increasing the likelihood of marginal banks failing for reasons unrelated to abuse, and increasing ultimate costs to the insurance fund. Although there can be many sorts of doubts about the effectiveness of government regulation, it seems unlikely today that any ineffectiveness would be due to lack of enforcement weapons.

C. Implications of Insider Abuse for a Special Bank Insolvency Regime

General bankruptcy law recognizes the possibility of insider abuse, and addresses the problem primarily through the law of fraudulent conveyances and preferences, and the law of equitable subordination. This bankruptcy law operates through standards, usually with a rebuttable presumption that a transaction was done in good faith. Those wishing to challenge the transaction must then come forward with evidence tending to show that insider abuse indeed occurred. This presumption fits well with the evidence that insider abuse is not an important cause of most industrial insolvencies.

If insider abuse is prevalent enough, however, then the default rule should shift. That shift may place the burden of proof on the insider to show that abuse did not occur. Or, the standard could move all the way to a rule conclusively presuming abuse. In corporate law, courts apply greater supervision to certain categories of transactions having a high risk of abuse. [FN185] At the extreme, such as where an agent makes use of a principal's property without authorization, courts apply a conclusive rule against the agent, and will not even inquire into the reasons for the use. [FN186]

The choice among default rules raises the familiar issues of choosing between rules and standards. [FN187] As applied to insolvency rules, three factors would seem especially relevant. First are the transaction costs of doing a case-by-case analysis under a standard. As these transaction costs rise, it becomes more desirable to set a firm rule. Second are error costs of finding insider abuse when it did not exist, or no insider abuse when it did. As these error costs rise, a standard becomes more desirable. Third, but perhaps most importantly here, is the likelihood that insider abuse contributed to the insolvency.

Even in the current period, when bank insolvencies can readily happen due to stiff competition, insider abuse probably plays a greater role in the failures of banks and other financial institutions than in failures of industrial corporations. Where a showing is made of the great prevalence or likelihood of abuse, there is a justification for special rules, in or out of insolvency, to deter or compensate for that abuse.

The systemic effects described in Part II, including the arguments for saving the fund or deferring to FDIC expertise, give an undifferentiated basis for agency victory over any competing claim. The "model of insider abuse" provides a more discriminating basis for special insolvency rules: Where there is a high enough likelihood that a transaction or failure was due to abuse, then the rules shift against the party who likely participated in the abuse.

The "model of insider abuse" thus provides a plausible policy basis for passage of many of the recent bank insolvency rules. As discussed below in Part V, both cross-guarantees and D'Oench powers can be explained as ways to address insider abuse, and cross-guarantees with more validity. Other rules linked to insider abuse include: strict rules concerning repudiation of contracts and leases, broadened definitions of fraudulent conveyance, asset freezes, FDIC discretion to adjust priority among creditors of the failed bank, unique agency enforcement powers, and priority of the FDIC over certain bankruptcy rules. [FN188] Any of these rules might be justified if a strong empirical basis is shown for believing that these transactions or insiders are especially likely to be involved with abuse. As discussed in the next Part, however, the political needs of the FDIC and Congress, rather than public policy analysis, likely created an excessive number of strict rules. Ironically, the presumption of insider abuse led to passage of new rules just as a far greater proportion of failures was caused by stiff competition.

IV. PUBLIC CHOICE AND SPECIAL BANK INSOLVENCY RULES

The Article up to this point has essentially explored whether the special regime in bank insolvency has evolved for good public policy reasons. Reducing the systemic effects of bank failures and preventing insider abuse are policy goals that can gather widespread support, and both goals can be used to justify as well as explain special rules in bank insolvency.

Public choice theory offers a quite different approach to explaining the special bank insolvency regime. Concerned with the rational analysis of the means and ends of political actors and institutions, public choice theory can help explain and predict political behavior. [FN189] Public choice theory can also assist in evaluation of public policies. In particular, when public choice offers a strong explanation for why an outcome is in the interest of some particular set of institutions, the result can be to cast doubt on whether the proffered public policy rationales are in fact being sought or achieved.

This Part presents a public choice analysis of the special bank insolvency regime. In short, the FDIC can be seen as an agency trying to expand its power and protect the insurance fund. Members of Congress, meanwhile, have been concerned since the passage of FIRREA with avoiding voter backlash for the bailout. With the interests of the FDIC and Congress thus aligned, the result has been a "strict" approach to bank and thrift failures, expressed as the grant of extraordinary new powers to the FDIC. The temptation for the agency and Congress is to push costs of the bailout off-budget and onto third parties. These incentives to be strict lead to the suspicion that the recent flurry of bank insolvency laws have been based more on politics than on accurate public policy.

A. The FDIC and the Public Choice of Bank Insolvency

Identifying the goals of the FDIC will help us to understand recent developments in bank insolvency law. A recent article by Ronald Cass and Clayton Gillette canvasses the public choice explanations of bureaucratic objectives. [FN190] Starting from William Niskanen's hypothesis that agencies maximize their budgets, [FN191] Cass and Gillette also examine such goals as furthering agency programs, providing service to clients, and advancing the careers of bureaucrats.

In understanding the recent behavior of the FDIC, I suggest that special attention should be paid to its unique goal of protecting the deposit insurance fund. Historically, the FDIC fund for insuring bank deposits grew each year from 1933 to 1988. [FN192] During this period, banks were the key FDIC constituency. Banks paid higher insurance premiums than the FDIC used in any year, and then received annual rebates. The FDIC could please the key banking constituency by assuring rebates and allowing the fund to grow slowly. After 1988, however, the rebates stopped. [FN193] As the insurance fund began to decline, taxpayers and especially the Congress became the key constituents. They began to fear, with justification, that general government revenues would be needed to protect the insurance fund. [FN194]

I propose that in the stressful period since 1988, when losses from bank failures became significant, a key goal of the FDIC has been to maximize the insurance fund. If this hypothesis is correct, then many agency actions should be correctly predicted or explained by reference to their tendency to maximize the fund. [FN195] In contrast to Niskanen's general model of public agencies, where there is no clear measure of agency output, [FN196] the amount in the deposit insurance fund serves as a public, objective measure of agency performance. [FN197] The agency would both try to increase the balance of the fund, and rebut any information suggesting that the fund was at risk. [FN198]

The FDIC has witnessed an unusually apt bureaucratic lesson about why it should be concerned with the size of the insurance fund. The failure of FSLIC to maintain its insurance fund resulted in that rarest of Washington events-the death penalty for an entire agency. As the thrift crisis deepened, a great deal of public attention focussed on the ever-larger projections of the FSLIC deficit. FHLBB Chairman M. Danny Wall suffered withering criticism for his failure to acknowledge the magnitude of the problem. [FN199] In FIRREA, the FHLBB was abolished, and its insurance fund was placed under the control of the FDIC. [FN200] Faced with the lessons of FSLIC, FDIC administrators and employees have an excellent incentive to protect today's insurance funds.

Recent actions by the FDIC seem to fit the hypothesis that the agency is seeking to maximize the insurance fund, rather than the traditional Niskanen hypothesis that the agency is seeking to maximize its budget. If the agency were trying to maximize its budget, then one would expect significant discussion about its tendency to hold on to properties rather than to dispose of them quickly. [FN201] But my own reading in the trade press has yielded no reason to think that the agency is trying to manage an enormous real estate empire in order to increase its own budget.

Other qualitative evidence supports the idea that the FDIC is seeking to maximize the insurance fund. FDIC Chairman William Seidman responded harshly to a 1990 report by three economists who estimated that the bank insurance fund was insolvent at that time on an actuarial basis. [FN202] In light of Seidman's political astuteness, the vehemence of his response is a good indication of the seriousness he attached to public perceptions of the fund's solvency. Furthermore, the FDIC has lobbied vigorously for discretion to set the level of assessments that the banks pay into the insurance fund. [FN203] Such discretion is important to agency calibration of how best to maximize the fund over time. Finally, and most broadly, the FDIC has adopted long-range litigation and lobbying strategies to maximize the flow of closed-bank assets into the fund. [FN204] These efforts have resulted, for instance, in expanded D'Oench and cross-guarantee powers. By contrast, there is little reason to think that the FDIC is holding on to failed bank assets, rather than selling them, in order to increase the size of its budget or the empire under its control.

Inter-agency politics has encouraged the FDIC to focus on protecting the insurance fund. At least until FDICIA in 1991, the Federal Reserve and the Office of the Comptroller of the Currency effectively protected their existing powers to regulate the important solvent banks. [FN205] The FDIC's power as insurer to control solvent banks has remained limited. [FN206] The chief outlet for the FDIC's legislative efforts has thus been the law of bank insolvency.

In pursuing its goal of maximizing the insurance fund, the FDIC has some systematic advantages in Congress. The bank and thrift industries may be at a particular disadvantage in lobbying for less strict rules in insolvency. In opposing a rule aimed at insider abuse, the bank and thrift industries have to explain why it is better to be "soft" on persons or institutions that are costing taxpayers large sums of money. The internal politics of the industries also weaken lobbying efforts on insolvency issues. The strong banks and thrifts place little priority on insolvency law, while the weak banks and thrifts have constrained resources and suffer from the suspicion that they are trying to escape the consequences of their own fraud or mismanagement. A related advantage for the FDIC is that it has been able to pursue its legislative goals repeatedly, in each of the banking statutes in recent years, while its natural opponents on insolvency law have been the shifting set of institutions that in a given year were strong enough to lobby but weak enough to be concerned about insolvency. [FN207]

The models developed by James Q. Wilson [FN208] and Michael Hayes [FN209] are helpful in understanding descriptively where the FDIC has been most likely to achieve its legislative goals. The FDIC has sought a variety of legislation that gives it "concentrated" benefits, that is, benefits that flow primarily or exclusively to the FDIC and are designed to achieve the major FDIC goal of maximizing the fund. Where the concentrated benefits to the FDIC are paired with "diffuse" costs to others, the FDIC success is greatest. Part V argues that the FDIC's expansion of the D'Oench powers has been aided by an extremely diffuse distribution of costs. By contrast, where the concentrated benefits to the FDIC are paired with concentrated costs to another group, the FDIC has encountered much stiffer opposition. The best example of this is the loss which the FDIC suffered in FIRREA on recovering on many large directors' and officers' insurance contracts. On this issue, the agency was unable to defeat a coalition of large directors' and officers' insurers and solvent banks that wanted the insurance. [FN210] From a normative standpoint, there is great reason for concern about a particular insolvency rule where the concentrated benefits to the FDIC are clear, but the political opposition diffuse and thus unlikely to serve as an effective check.

B. Congress and the Public Choice of Bank Insolvency

The above discussion of the FDIC's legislative goals and achievements draws on the interest group branch of public choice theory, which analyzes the factors contributing to the legislative success or failure of organized political entities. A complementary branch of public choice theory focuses on the goals and actions of the legislators themselves, and usually emphasizes the legislators' goal of getting reelected. [FN211]

A recent book by R. Douglas Arnold, The Logic of Congressional Action, [FN212] provides an excellent framework for explaining how members of Congress have acted with respect to bank and thrift insolvency. Arnold stresses that in seeking reelection representatives predict the "potential preferences" of their constituents on issues that may become the subject of future political campaigns. Representatives are especially concerned with the subset of issues that are "traceable" to the incumbents, and adopt strategies to minimize the risk that these traceable issues will be contrary to the preferences that voters will have come Election Day.

Arnold persuasively sets forth three conditions that must be met to make an effect traceable to a representative, and thus the basis for a vote against the incumbent: "a perceptible effect, an identifiable governmental action, and a legislator's visible contribution." [FN213] Bank and thrift insolvencies meet these criteria precisely. The "perceptible effect" for voters would be the enormous cost of the bailout, which has been reported by the press to be between $100 billion and $1 trillion. Individual voters might readily interpret such sums as a significant personal burden, expressed as increased future taxes or higher costs for servicing the budget deficit. The "identifiable government action" would be the repeated bills passed by Congress to fund the bailout, coupled with earlier legislation deregulating the thrifts and arguably causing the crisis. The "legislator's visible contribution" is the roll call vote by the incumbent in favor of providing money to resolve the insolvencies. Some members have made even more visible contributions, such as heading a banking committee or making headlines in a thrift-related scandal.

Recent elections show that voters have indeed become disposed to punish their incumbent representatives for the bailout. In the House, several members seem to have been defeated on the thrift issue in 1990, and many more were attacked on it during campaigns. [FN214] In the Senate, the Keating scandal caused great embarrassment for five Senators, and contributed to Senator Alan Cranston's decision not to seek reelection. [FN215]

Members of Congress have responded to this traceable issue with strategies that respond to each of the three conditions identified by Arnold. [FN216] Congress can combat the "perceptible effect" of the bailout cost by adopting what might be called a "strict" strategy for reducing the bailout cost. A large number of recent enactments fit within the "strict" strategy. [FN217] For instance, FIRREA raised fines for bank-related violations to up to $1 million per day, and numerous insolvency provisions have made it easier for the agencies to recover assets that arguably belonged to the insured institution, and thus to the insurance fund. [FN218] The 1991 amendments calling for mandatory early closure of weak banks also demonstrates a strict approach for reducing future losses. [FN219] In addition, Congress's decision to split the cost of the bailout into several separate bills may aim to reduce the "perceptible effect" of the bailout-voters might remember the most recent $25 billion authorization rather than aggregate the full cost of hundreds of billions of dollars.

Congress can try to avoid blame for "identifiable government action" either by justifying the government actions or by shifting the blame to others. Members can articulate a public policy basis for the bailout costs, namely that deposit insurance exists to protect the money of ordinary depositors. The strategy is to maintain the still-popular system of deposit insurance, while demonstrating a "strict" approach of punishing thrift kingpins.

Shifting blame onto the President and the agencies can also reduce the chance that voters will identify incumbents as worthy of electoral punishment. There has been no shortage of hearings criticizing bank and thrift regulators for causing bank and thrift problems. [FN220] The Senate's 1991 refusal to confirm the renomination of Robert Clarke as Comptroller of the Currency [FN221] is a clear example of blaming the President and the agencies. Clarke was the senior bank regulator directly responsible to the President, and Clarke was severely criticized by Democrats, led by Senator Donald Riegle, in part for laxness in the regulation of national banks. [FN222] Indeed, such blame-shifting may explain the more general outlines of recent insolvency law-Congress can give broad powers to the FDIC and RTC, and then criticize the agencies for their inevitable failure to solve bank and thrift problems.

Arnold's third condition of a traceable issue, a "legislator's visible contribution," has also been the subject of defensive action by Congress. The strict strategy has allowed incumbents to show visible support for being "tough on S&L kingpins." [FN223] Particular legislators who were vulnerable to charges of being responsible for the thrift crisis have taken vigorous steps to show their toughness on the issue. [FN224] Most striking is the escalating reluctance of Congress to authorize funding for the bailout. The House approved the November 1991 authorization of $25 billion for the RTC on a voice vote, apparently to avoid members having to go on record as supporting the bailout. [FN225] When the RTC's temporary authorization to close thrifts expired in April 1992, the House completely refused to vote for continued spending, despite estimated costs of delay ranging up to $1.4 billion. [FN226] Such unwillingness to allow needed spending is a remarkable sign of legislators' avoiding a "visible contribution" to the thrift bailout. [FN227]

C. Summary

Congress's strict strategy has beautifully complemented the FDIC's institutional desires to increase agency turf and protect the insurance fund. [FN228] Both the FDIC and Congress prefer to have third parties shoulder the costs, rather than pay from the insurance funds. The list of third-party payors has become impressive: state and local governments deprived of taxes; borrowers and others affected by D'Oench; almost any bank insider with deep pockets; lawyers and accountants unlucky enough to be near the scene of a failure; and so on. In light of the public choice conclusions, the question then becomes how to assess normatively the strict system that has developed.

V. APPLICATION TO CURRENT BANK INSOLVENCY PROBLEMS

The previous Parts have developed three accounts that explain the development of special bank insolvency law: the public policy stories based on the systemic effects of bank failure and the problems of insider abuse, and the public choice story of the goals of Congress and the FDIC. These three accounts show how a special confluence of public policy and public choice occurred in 1989, when the decisive statutory shift occurred in FIRREA. FIRREA came at the time of the first really substantial taxpayer funding of the thrift insurance fund, [FN229] and at a new height of publicity about the exploits of Charles Keating and other thrift insiders. Meanwhile, the bank insurance fund seemed solvent, with its major losses readily explicable. [FN230] In this setting, the FDIC and its popular Chairman, William Seidman, became an attractive "white knight" for solving the thrift crisis. The FDIC took over the thrift insurance fund and became manager of the RTC's cleanup efforts. To do this tough job, the FDIC sought and received broad special powers in insolvency, increasing its turf and gaining tools for protecting the insurance fund. Congress could feel comfortable with its actions in the insolvency area: FIRREA was consistent with strong public policy arguments; it was "strict" on thrift kingpins (helping politically on a "traceable" political issue); and it delegated power to the best available agency (accomplishing policy goals while reserving Congress's ability to blame the agency if things did not work out).

The three accounts developed in this Article thus combine to give a good account of how the FDIC has gained its extraordinary range of special powers in bank insolvency. All the major forces have pushed in the same direction, toward a greater divergence of pre- and post-insolvency law. A remaining task is assessing whether the current rules are normatively desirable. I conclude that there is often reason for concern that current superpowers are an overreaction to the problems that justified some special rules in bank insolvency. When all three accounts push the same way, where are the brakes? The danger of overreaction is greatest where there are no effective political checks on the incentives of Congress and the FDIC to be strict.

In this Part, particular aspects of bank insolvency law are first examined, to understand how the three accounts together can justify or explain two notable areas of bank insolvency law-FIRREA's cross-guarantees provision and the D'Oench powers. Second, a more global assessment is made, focusing on the costly consequences of having too large a gap between pre- and post- insolvency banking law. The most important consequence is what I term the "black hole" effect, in which weak but solvent banks may be pulled inexorably down into insolvency.

A. Assessing and Explaining Particular Aspects of Bank Insolvency Law

1. Cross-Guarantees. For a normative assessment of the cross-guarantees provision, a key question is whether there is enough risk of insider abuse to justify the extraordinary FDIC powers. The systemic account is only weakly related to cross-guarantees, much as any rule that helps save the insurance fund. From the public choice perspective, passage of cross-guarantees is consistent with the goals of Congress and the FDIC, while the banks on whom the costs are concentrated seem to retain some potential to check agency overreaching.

a. Cross-guarantees and the "model of insider abuse." As described above, [FN231] FIRREA's cross-guarantees provision allows the FDIC to charge affiliated banks and thrifts (but not the holding company or non-bank affiliates) for the costs of a failure, and subordinates the claims of shareholders and non-bank affiliates to the FDIC claim. This provision arose from the FDIC's experience with large Texas bank holding companies, including MCorp and Texas American Bankshares. These holding companies had set up numerous separately incorporated banks in order to comply with strict state laws against branching. In the FDIC's opinion, these holding companies employed a strategy of dumping liabilities into a few of their banks, which became deeply insolvent, while retaining assets and shareholder worth in other banks. When some banks neared insolvency, the holding companies and solvent banks refused regulators' demands to support the weak banks. [FN232] The FDIC and the Federal Reserve then attempted novel legal means, with limited success, for gaining the value in the holding company and solvent affiliates. [FN233] Meanwhile, the FDIC sought and received from Congress the cross-guarantees powers, to avoid similar problems in the future.

In assessing cross-guarantees, first imagine, consistent with the "model of insider abuse," that many bank failure were accompanied by systematic transfer of assets to an affiliated, solvent bank. Bank owners could use deposit insurance as a convenient way to dump losses, while retaining all profits. Would cross-guarantees be desirable in such circumstances? Probably so. Consider the alternatives. [FN234] Relying on fraudulent conveyance law would likely fail to capture much of the fraud and would have enormous transaction costs. Routine transactions between affiliated banks can be incredibly numerous, especially in the Texas model where one or more "lead banks" serve as a conduit for most loans or other investment in assets outside of the holding company. Such a corporate structure makes it terribly difficult to distinguish a routine transaction from part of a fraudulent scheme.

More promising would be an FDIC power to take advantage of the doctrine of equitable subordination, as creditors can do in bankruptcy. Under this approach, the FDIC would prove in federal court that a pattern of insider domination should justify subordinating the insiders' claims to those of bona fide outside creditors. [FN235] Robert Clark has shown that equitable subordination is a desirable substitute for fraudulent conveyance law in highly complex cases where it would be costly or impossible to make findings as to the fraudulent nature of each transaction. [FN236] The chief advantage of equitable subordination, compared with cross-guarantees, would be remedial precision-insider claims would be subordinated only to the extent an abusive scheme was shown, and would not be presumed except where abuse is evident. A second advantage would be the increased likelihood of debt financing to the troubled bank by holding companies and non-bank affiliates, thereby reducing the likelihood of insolvency and costs to the FDIC. [FN237]

In a world of pervasive abuse, these advantages would have little weight. First, a rule assuming abuse would have greater accuracy and lower transaction costs, and thus be preferable to a standard where the creditor has to come forward with evidence of abuse. Second, funding by other insiders would deserve equity treatment under the rules of equitable subordination, and cross-guarantees would again have the advantage of simplicity and lower transaction costs. As abuse becomes less pervasive, these same arguments would apply, but with lesser weight. The key normative question would be whether abuse was pervasive enough to justify the strict rule.

Even assuming that abuse is less pervasive, an automatic rule such as cross- guarantees may be justified as a means to prevent abuse from occurring. [FN238] The bright-line nature of cross-guarantees may make it relatively hard to evade the purpose of the rule and dump liabilities onto the fund while preserving value to shareholders. [FN239] If the rule is difficult to evade, then it will seldom have to be used to address fraud; potential wrongdoers will realize the pointlessness of fraudulently transferring assets to affiliated banks. [FN240]

b. Cross-guarantees and systemic effects. On first impression, the mere existence of deposit insurance would seem to give little reason for special rules for handling transfers with affiliates. To the extent that cross- guarantees allow the FDIC to grab affiliate assets in the event of an insolvency, there is the general temptation to save the fund by favoring the FDIC. But something more should be shown in order to favor a rule, especially in light of the possibility that cross-guarantees may discourage debt financing of the banks and thus ultimately increase the likelihood of insolvencies and the cost to the fund.

A somewhat stronger systemic argument focuses on how the FDIC is to sell the insolvent bank. The clear priority rules embodied in the cross- guarantees allow an acquiror to price a bank or its assets quickly, without speculating on the nature of the claims of bank affiliates or the likelihood of creditors eventually making out a case for equitable subordination. A rule fostering speedy sale helps the fund, because the value of bank assets deteriorates steadily the longer they are held by the government receiver. [FN241] Because of the special problems the government seems to experience in maintaining the value of a receivership, a special rule fostering speedy sale may be justified. [FN242]

A related systemic argument looks to the going concern value of the affiliated banks. The FDIC may wish to sell all affiliated banks as a package, in order to preserve the business advantages that existed from their joint operation. Precisely this rationale was given for using cross-guarantees against the Maine affiliate of the Bank of New England. [FN243] Preserving going concern value may well be more a more efficient use of the banking assets than splitting control between the FDIC and the holding company. This efficiency point is bolstered by the usual equitable arguments that favor saving the fund, rather than leaving value in the hands of the holding company that allowed the bank to fail.

c. Cross-guarantees and public choice. The cross-guarantees power fits well into the public choice perspective developed in Part IV. The new powers furthered the FDIC's goals of increasing its turf and protecting the deposit insurance fund. Its turf was increased by the strong new power, coupled with the ability to waive the cross-guarantees at the agency's discretion. The insurance fund was protected by the FDIC ability to take assets from affiliated banks, and by the prophylactic effect of preventing future sequestering of assets. The cross-guarantees provision also met Congress's goal of protecting itself on a traceable issue. Members could now proudly show that they "did something" to prevent abuses of the fund as allegedly occurred in MCorp.

A modest puzzle is why the banking industry failed to defeat the cross- guarantees provision. The costs of the provision would seem to be concentrated fairly directly on bank shareholders, who now lose value when any bank affiliate fails, and bank officers, who will lose their jobs when an affiliate's insolvency triggers their own bank's insolvency. [FN244] Why wasn't this provision defeated by industry?

Several factors explain the political inability of the banks to defeat cross- guarantees. The bank lobby lacked strong allies on this issue. [FN245] Publicity surrounding MCorp and other Texas insolvencies made a solid political and policy case for a prophylactic rule such as cross-guarantees. In addition, the major burden of the provision occurs only when a bank becomes insolvent; and, as explained above, [FN246] banks concerned about insolvency are in an especially weak position on both policy and public choice grounds. Finally, the FDIC, during consideration of FIRREA, agreed to material amendments to the cross-guarantees proposal in ways that addressed the concerns of solvent banks while meeting FDIC goals. [FN247]

d. Assessing cross-guarantees. Application of the three accounts developed in this Article suggest that there is a reasonably strong normative basis for the cross-guarantees power of the FDIC. The "model of insider abuse" suggests that it is especially easy to shift value among affiliated banks [FN248] but very difficult to unravel the transactions in time to facilitate the resale of the insolvent bank and its assets. If these transactions cannot be quickly unraveled through either the law of fraudulent conveyance or equitable subordination, then systemic concerns suggest the desirability of the FDIC having power to sell the affiliated banks together to retain going concern value. Equity also favors protecting the insurance fund compared with protecting shareholders or affiliate shareholders who may have contributed to the insolvency. From a public choice perspective, the existence of a concentrated group that absorbs the cost of the cross-guarantees provision suggests a potentially important political check on excesses in the area. [FN249] The combination of plausible policy rationales and potential political checks on the FDIC gives a fairly strong justification for current cross- guarantees law.

2. D'Oench Powers. The FDIC and thrift agencies have invoked both systemic effects and insider abuse as justifications for expansive D'Oench powers. Although these justifications seem strong for the core example of regulators defrauded by secret agreements, D'Oench powers have been extended to new cases where the justifications are less strong, and where the agencies have not yet clearly come forward with supporting rationales. [FN250]

a. D'Oench powers and systemic effects. Banking agencies have invoked a number of systemic reasons for broad D'Oench powers. Most importantly, the entire system of deposit insurance depends on the ability of regulators to assess a bank's books. Strict rules against unwritten agreements can thus be justified in order to allow regulators to supervise lawful behavior in solvent institutions and to decide when to declare insolvency. In turn, accurate decisionmaking by regulators should reduce costs to the insurance fund.

Institutional factors also argue for strong D'Oench powers. D'Oench powers are similar to cross-guarantees in that they help the FDIC speedily determine the value of an insolvent bank. The FDIC has argued that this speedy determination helps it meet its statutory obligation of choosing the least-cost means of resolving the institution. [FN251] There is also an unusual difficulty in getting reliable testimony from the bank employee, often an officer, who took action prior to insolvency that forms the basis for the third party's claim or defense. [FN252] On the one hand, the employee has typically been fired, has no particular loyalty to the receiver, and may have more loyalty to the other party to the agreement. Therefore, the employee may not testify as helpfully to the bank as would a current employee. On the other hand, the employee may be facing an enforcement proceeding for actions that contributed to the insolvency. Therefore, the employee may have incentive to testify even more helpfully to the government than truth would warrant. In light of these conflicting impulses, a clear D'Oench rule preventing the use of such testimony may be desirable.

The strength of these systemic arguments, however, is far from clear. As to the reliability of employee testimony, the legal system faces innumerable factual settings where witnesses have incentives to distort the truth. Nevertheless, testing truth through cross-examination is usually considered superior to excluding entire categories of claims from the courtroom (although more nuanced rules could be developed where the risk of inaccurate testimony is greatest). And the fears of inaccurate testimony must be great to justify defeating otherwise valid claims of third parties, individual claims which range into the millions of dollars. [FN253] Moreover, similar incentives can exist for former employees of an ordinary corporation in bankruptcy, yet there is no similar rule barring oral claims and defenses in bankruptcy.

The systemic advantages of speedy bank valuation are even more dubious. The available empirical evidence suggests that D'Oench factors have not historically been important to the agencies' decision of whether to liquidate. [FN254] Indeed, it is difficult to understand how the agency could make a loan-by-loan assessment of a bank's balance sheet in time to make its determination as to the least-cost means of resolution. [FN255] Furthermore, contractual means are well known for selling a bank quickly even when uncertainty exists as to its value, so that D'Oench powers do not materially speed the ability of the agency to sell the institution. [FN256]

Even if D'Oench powers do not clearly offer systemic advantages, they may be justified as doing little or no harm in the long term. Persons doing business with banks in the future should conform their conduct to the D'Oench rules. For instance, borrowers will insist that contracts and contract modifications meet the strict conditions of section 1823(e). To the extent that borrowers cannot control how the bank keeps its written records, they will place a risk premium on transactions with banks. [FN257]

This story of the market smoothly adjusting may be too rosy, however. From an equitable point of view, agency superpowers unfairly harm those who are ignorant of the intricacies of D'Oench doctrine, such as consumer borrowers. The superpowers are unfair during the transitional period when bank insolvency is becoming more common and D'Oench powers more expansive-good- faith transactions will be vitiated by D'Oench powers, and bank receivers will recover more value than the amount for which the bank had originally bargained. From an efficiency point of view, expan