Christine E. Blair^*

Introduction

The issue of whether, or to what extent, banking and commerce should be allowed to mix is again the focus of a public policy debate. The issue often arises when criteria for permissible activities for a bank and its owners, subsidiaries, and affiliates are being discussed, as they are now. Although there is no hard evidence that combinations of banking and commerce are harmful, there is no evidence that they are beneficial, either. Nevertheless, developments in the foreign and domestic marketplaces suggest that combinations involving banking and commerce are becoming more common. Thus, the debate has been renewed.

The current debate centers on industrial loan companies (ILCs), also known as industrial banks. ILCs are state-chartered institutions that have banking powers, subject to certain restrictions on lending and deposit taking. ILCs are regulated by their state chartering authorities and, at the federal level, by the FDIC. The unique nature of the ILC charter has kept these institutions outside the purview of the Bank Holding Company Act (BHCA). As a result, the parent companies of ILCs include a diverse group of financial and commercial firms.¹

Two pieces of legislation passed by the U.S. House of Representatives in 2003 have focused attention on whether ILCs should be considered equal to other insured depository institutions with regard to powers such as interstate banking and payment of interest on business checking accounts.² Consumer groups and community bankers have responded to the proposed legislation by raising questions about its competitive effects. In particular, concerns focus on the possibility that commercial entities, which in certain states can enter banking by acquiring an ILC charter, could branch nationwide. For example, in 2002, Wal-Mart attempted to acquire an existing ILC in California. In response, the California legislature amended the state's law, thereby prohibiting a commercial entity from making such an acquisition, and Wal-Mart subsequently withdrew its notice. Other concerns focus on whether federally insured depository institutions, including insured ILCs, should be allowed to pay interest on business checking accounts; some people argue that if they were, the ILCs would have a competitive advantage over other insured depository institutions. Finally, fears have been expressed that the failure to prohibit such a mixing of banking and commerce could lead to a situation like that in Japan, where informal links between commercial firms and banks have raised safety-and-soundness concerns.

The Federal Reserve Board has expressed concern that expanding the powers of ILCs would weaken the legal barriers separating banking and commerce. The Board argues that there is a long-standing policy of separating banking and commerce and that the proposed expansion would undermine that policy. Although the FDIC has the authority to examine the parent of any ILC, the Federal Reserve Board argues that the absence of federal oversight of the owners of ILCs threatens the safety and soundness of the banking system.³

As the primary federal regulator of ILCs, the FDIC has expressed the view that these institutions pose no greater safety-and-soundness risk than do institutions with any other charter.⁴ Rather, the challenge facing bank regulators is to ensure that market innovation can take place while maintaining the public's confidence in the banking system. As FDIC Chairman Donald Powell has noted, regulators must guard "against the possibility that the regulatory system itself does not impair the vital process of innovation and change that is the lifeblood of the American marketplace."⁵

And so the stage is set and the debate over banking and commerce continues.⁶ The relevant questions are should banking and commerce should be allowed to mix, and if they mix, should the combination be regulated? This paper examines the arguments in terms of the public interest; reviews the evidence about an alleged long-standing principle of separation; explores the benefits of, and then the risks posed by, affiliations between banking and commerce; discusses firewalls and prudential supervision; spells out two approaches to regulating affiliations; and concludes with a summary and a discussion of policy implications.

Separation versus Affiliation: The Public Interest

There are generally two views on whether banking and commerce should be separated. The first view argues that a line of separation must be maintained because the risks of allowing banking and commerce to mix outweigh the possible benefits. The failure to maintain a separation of banking and commerce, especially in terms of ownership and control of banking organizations, could have potentially serious consequences, ranging from conflicts of interest and the lack of impartiality in the credit decision-making process to the unintended expansion of the financial safety net. To adequately protect the insured entity from such abuses (it is argued), the insured entities' corporate owners need to be subject to federal supervision and regulation.⁷ This viewpoint has been articulated over the years by (among others) the Federal Reserve Board, some members of Congress, and community bankers; many of these advocates of separation claim that their position is based on a long-standing principle of the separation of banking from commerce (this claim is examined in detail below).⁸

The other view argues that mandating a separation of banking and commerce prevents the benefits of affiliation from being realized and can result in an inefficient allocation of resources. Given adequate supervisory oversight of the insured entity, federal regulatory and supervisory authority over the corporate owners of the insured entity represents an unwarranted hampering of the market process that is neither necessary nor desirable. This view has been expressed by (among others) the FDIC, some members of Congress, and public policy groups.⁹ The FDIC has long argued that with certain safeguards in place to protect the bank and ensure the safety and soundness of the banking system, affiliations between banking and commerce should be permitted.

Although the current debate centers on the industrial loan charter, the underlying policy issues—which have been debated for many years—come down to whether the public interest is served when affiliations between banks and commercial entities are prohibited.

Testifying before Congress on financial services reform in 1987, the FDIC's then-chairman L. William Seidman argued that the public interest would be best served by a financial services industry that met four objectives: the financial system should be viable and competitive, the banking system should be operated in a safe and sound manner, customers should realize benefits from enhanced competition, and the system should be flexible enough to respond to technological change.¹⁰ Consistent with these objectives, the regulatory and supervisory structure of banking should be the simplest and least costly one available.

The question facing policy makers then was—and continues to be—whether these objectives can be met without restricting the ability of banks to choose the corporate structure that best suits their business needs. As Seidman noted: "The pivotal question . . . is: Can a bank be insulated from those who might misuse or abuse it? Is it possible to create a supervisory wall around banks that insulates them and makes them safe and sound, even from their owners, affiliates and subsidiaries?"¹¹ If so, then the banking and commerce debate should focus on how affiliations should be regulated so that the public interest is met.

A Long-Standing Principle of Separation?

The literature on the issue of a long-standing principle of separating banking and commerce is extensive.¹² This literature shows that the extent to which banking and commerce have mixed or have remained separate has been a function of the demands of the marketplace, the level of technology and the state of development of organizational and business structures. Recently, Haubrich and Santos (2003) dispel any notion that a separation of banking and commerce has been a long-standing principle in American banking history. They conclude that despite the regulations and prohibitions on certain activities and forms of control, extensive links between banking and commerce have existed and continue to exist and have often been facilitated by the use of arrangements very similar to those that have been prohibited by law.

For example, certain charter types-including limited—purpose consumer banks and ILCs—permit a mixing of banking and commerce. These charter types do not fit the definition of a bank under the BHCA and technically are not banks; in certain states, they can be owned by commercial firms. These firms, in turn, are not subject to the BHCA and are not required to become bank holding companies.¹³

And there is other evidence of banks exercising control over commercial firms, and commercial firms exercising control over banks, through various means. Sometimes, as legal restrictions were placed on the mixing of banking and commerce, certain exceptions were made that allowed commercial firms to retain their affiliations with banks. Examples include the limited number of nonbank banks that were grandfathered by the Competitive Equality and Banking Act of 1987 (CEBA), and the unitary thrift holding companies that were grandfathered under the Gramm-Leach-Bliley Act of 1999 (GLB). Sometimes, the mixing has resulted from the equity investments of banks, including investments in small business investment companies, equity acquired in loan workouts and equity kickers, and merchant banking activities. Outside of chartered banking, captive finance companies of large commercial firms (e.g., GE Capital) also approximate a mixing of banking and commerce. Moreover, individuals are permitted to hold a controlling interest in both a bank and a nonbank commercial firm. For example, in the case of chain banking organizations, federal regulatory oversight does not extend to the owner.¹⁴

International comparisons of the treatment of banking and commerce are instructive. The U.S. practice of prohibiting affiliations between banking and commerce contrasts with the practice of most other industrialized countries, since in these countries linkages among banking and commercial entities in the form of ownership and control are common. Throughout Europe, where universal banking is common, and in Japan, where the keiretsu is a dominant business form, banking and commerce traditionally have had greater freedom to mix.¹⁵

U.S. banking is in fact characterized less by a tradition of being separate from commerce than by regulatory attempts to separate it from commerce. Since the banking crisis and economic depression of the 1930s, these attempts have focused on prohibiting affiliations between banking and commercial firms—that is, on separating banking from commerce at the ownership level. In 1933, responding to the general belief that the nation's banking and economic problems had been caused by conflicts of interest between banks and their securities affiliates, Congress passed the Glass-Steagall Act, which prohibited affiliations between commercial and investment banking. Two decades later, in 1956, a general and long-standing distrust of large banking conglomerates combined with the increased merger activity of the 1940s and early 1950s led to the passage of the BHCA, which separated banking from commerce by restricting the activities of owners and affiliates of banks. The BHCA defined bank holding companies and established a framework for their regulation by the Federal Reserve. The restrictions on ownership and affiliation that are currently in effect stem from the BHCA and its subsequent amendments.

Throughout the remainder of the twentieth century, rapidly changing technology and the changing nature of banking led to increasing demands for the banking system to be restructured and given broader powers.¹⁶ At the same time, the regulatory line separating banking from commerce was being weakened as banks increasingly found ways to engage in a range of financial activities.¹⁷ And other financial services providers found ways to offer bank-like products to their customers, one example being the cash management account offered by securities firms. After repeated congressional attempts to address financial modernization, GLB was passed in 1999, effectively acknowledging and extending the degree to which banking organizations were permitted to engage in nonbank financial activities.¹⁸

In summary, banking has never been absolutely separate from commerce. Although the activities permitted to banks have always been subject to prohibition, restrictions on affiliations with nonbank firms are relatively recent. Moreover, despite the regulatory line prohibiting affiliations between banking and commercial firms, it is likely that the market will continue to move toward a greater blending of banking and commerce. The linkages that exist between banking and commerce outside of the current restrictions on ownership or activity can also be expected to continue.¹⁹ Thus, as has recently been noted, "It is perhaps better to replace the claim that banking has been separated from commerce in the United States with the observation that regulations have attempted to separate banking and commerce."²⁰

The Potential Benefits of Affiliation

The potential benefits of mixing banking and commerce are well known and have been discussed in the economics literature.²¹ Among the potential benefits are operational efficiencies—cost and revenue synergies—including economies of scale, economies of scope, and informational efficiencies. Other potential benefits may result from greater product and geographic diversification, access to new sources of capital, and enhancement of the global competitiveness of U.S. banks.

Cost synergies can result from economies of scale (when increasing the scale of operations lowers the average costs of production) or from economies of scope (when costs of production are lowered by the production of products that share inputs). Finding empirical evidence for the existence of these economies in banking has proven to be difficult.²² However, the lack of demonstrable economies in banking does not imply a lack of cost complementarities between banking and other commercial activities. For example, the entrance of commercial firms into bank-like activities may be evidence of economies of scope. Technological innovations in recent years have made combinations of banking and commerce in the United States economically feasible and profitable. Changes in the cost structures of banks and commercial firms alike, which are the result of improvements in technology, also leave room for potential economics of scope.

Should economies of scope exist, they would provide one incentive for banks and commercial firms to seek mergers with one another. However, even though GLB has lessened the restrictions on affiliations among banks and securities and insurance firms, the limitations on bank activities and commercial affiliations have largely kept U.S. banks from availing themselves of the possible synergies.

A further incentive for affiliating may come from informational efficiencies.²³ For example, banks could have an incentive to hold equity positions in commercial firms if doing so would make it easier for them to gain the information necessary for their role as intermediary. In addition, holding equity can limit the bank's exposure to moral hazard. Bank financing of start-up ventures, when an equity claim substitutes for collateral, is an example. The equity claim can provide the bank with information about, and the ability to exercise control over, the commercial firm. These informational incentives would probably result in bank ownership of commercial firms, but not commercial ownership of banks.

Other literature has focused on the implications of banks holding equity positions in borrowing firms. For example, Pozdena (1991) cites arguments in favor of lifting existing restrictions on commercial and bank affiliations, noting that the ability simultaneously to hold the equity of commercial firms and lend to them is important to the successful intermediation of risky credits. Santos (1999) examines the implications—given deposit insurance—of equity stakes when funding is provided by a bank rather than a financier: mixed debt and equity are shown to control moral hazard. Haubrich and Santos (1999) argue that there is a liquidity synergy that gives banks an incentive to own a nonfinancial firm: by creating an internal market, merging with a nonfinancial firm increases the bank's efficiency in disposing of assets that back defaulted loans.

Other incentives for affiliations between banking and commercial firms include enhanced product and geographic diversification and greater access to capital.²⁴ Affiliations could lead to the diversification of the combined organization's portfolio risk, although the effect is likely to differ among banks of different size. Large banks with overseas operations that are permitted equity investments would probably see a smaller effect from affiliations at home than would smaller banks with no overseas presence. Although access to new capital was thought to be a compelling argument for affiliation in the early 1990s, when the cost of capital to banks was high, the immediacy of the need disappeared with the decade-long banking recovery and with structural changes (such as interstate branching) that have facilitated mergers.²⁵

Affiliations between banking and commercial firms can also enhance the global competitiveness of U.S. banks. As noted, many other countries do not place similar restrictions on the affiliation of banks with commercial entities, with the result that combinations of industrial, commercial and banking firms are common.²⁶ It can be argued that this potential benefit was tacitly acknowledged by the provisions of GLB that allow financial holding companies to engage in limited merchant banking activities for investment purposes. As such, these provisions result in a mixing of banking and commerce that was heretofore prohibited to U.S. banks.²⁷

Questions remain about the extent to which an incentive exists for banks and other firms—financial and nonfinancial alike—to affiliate. For example, large firms such as Sears Roebuck and Ford Motor Company took advantage of the unitary thrift charter during the 1980s, only to sell those thrifts subsequently. Krainer (2000) reports speculation that these firms may have wanted to capture tax losses at troubled thrifts rather than establish a long-term presence in banking. Another possible conclusion is that the combination of banking and commerce may be less attractive to commercial firms than some might expect.²⁸

The Potential Risks of Affiliation

The potential risks from allowing banking and commerce to mix that are cited most often are the potential for conflicts of interest and for the misallocation of credit; the fear of, or aversion to, a concentration of power—financial or economic—that could lead to monopolies; and the potential for an expansion of the federal safety net, which could expose the taxpayer to losses. (A discussion of these risks and whether they justify a separation of banking and commerce is presented in this section. The following section expands on the ways of managing the risks.)

Conflicts of Interest
Potential conflicts of interest exist whenever an entity that serves more than one interest is in a position to favor one of those interests over the other(s).²⁹ In banking, the opportunity for self-dealing at the expense of bank clients, beneficiaries of the bank's trust accounts, or bank creditors may create conflicts of interest. Conflicts of interest may also result from transactions between the bank and its affiliates, and these are the situations focused on by the debate about the separation of banking and commerce. For example, a bank affiliated with a commercial firm may choose to deny loans to the affiliate's competitors, may choose to lend preferentially to its commercial affiliate(s), or may illegally tie loans to purchases of the affiliate's products. Other sources of potential conflict of interest exist as well and are also discussed in this section.³⁰

Denying Credit to the Affiliate's Competitors. Does a bank with a commercial affiliate have an incentive to deny credit to its affiliate's competitors? From an economic perspective, given a competitive market for loans, a bank that unreasonably prefers its own affiliates is likely to suffer diminished earnings. When alternative sources of funds are readily available, the competitor will receive its funding elsewhere, but the bank will lose the profit it would have made on the loan. By denying loans to an affiliate's competitors, the profits of the consolidated banking organization will be lower than they would have been otherwise. However, such an incentive could exist if markets were not competitive, or if the affiliation yielded informational synergies so that the bank had a cost advantage over its competitors. In either case, the bank would have an incentive to deny credit to its affiliate's competitors.³¹ The question then becomes how to counteract the bank's incentive, which is discussed below.

Preferentially Funding the Affiliate. A situation in which the bank could choose to lend preferentially to its commercial affiliates, whether willingly or under duress, could arise because of the bank's access to lower-cost funds as a by-product of federal deposit insurance. This argument for separation, too, fails to hold in a competitive market for bank loans. Again, the nonaffiliated customer will be able to obtain loans from other providers at a competitive rate, and the bank's decision not to lend to its affiliate's competitors at competitive rates will result in lower profits. Moreover, this potential conflict has been addressed by Sections 23A and 23B of the Federal Reserve Act, which restrict the amount and terms under which banks can lend to their affiliates.³²

Tying Loans to Purchases of the Affiliate's Products. Tying loans to other business has the potential to harm the corporate customer or the bank. Tie-in arrangements are illegal under antitrust laws for all businesses, but Congress made it much easier to prove a tying arrangement when a bank was involved. The BHCA eliminates the need for the plaintiff to establish the economic power of the bank and the specific anticompetitive effects of the tie-in arrangement, as would be required under the general antitrust laws.³³ Section 106(b) of the BHCA prohibits anticompetitive tying practices: banks are prohibited from requiring customers to obtain nontraditional banking services or products in return for loans or a discounted banking service. This provision also precludes a bank from tying its banking services or products to the requirement that the customer not obtain some product or service from a competitor of the bank or its affiliates. Section 23B of the Federal Reserve Act, which requires transactions between affiliates to be at arm's length and on market terms, also serves to prohibit certain tying arrangements.

Tying is legally permissible in certain circumstances.34 For example, a bank may restrict the availability or vary the price of a loan on the condition that the customer also obtain a traditional bank product from the bank or an affiliate.³⁵ Tying violations generally involve the tying of loans with securities or insurance services or products. However, restrictions on tying can be avoided. For example, there is no violation if the loan is booked through a nonbank affiliate or parent holding company. Tying prohibitions do not apply to investment banks or to U.S. banks' business with non-U.S. companies. Moreover, tying is permitted when the transaction is requested by the customer rather than initiated by the bank.³⁶

The federal banking regulators have recently addressed the incidence and effect of illegal tying. In August 2003, the Federal Reserve Board issued a proposed regulation that would define the terms under which banks can legally engage in tying. The Office of the Comptroller of the Currency (OCC) also reported on the extent to which illegal tying poses concerns. Both agencies concluded that illegal tying was not a widespread problem among U.S. banks and that banks generally had adequate procedures to comply with antitying restrictions. In October 2003, the U.S. General Accounting Office released a study on the incidence of illegal tying and concluded that the extent of such tying is minimal and does not pose significant problems.³⁷

Other Potential Conflicts of Interest. Another potential conflict of interest is to use inside information to benefit the bank at the expense of a nonbank affiliate or an investor. The bank has access to private information as part of its commercial lending and trust activities, and typically the privacy of such information is achieved through the use of firewalls between respective departments of the bank. This conflict can also arise between the bank and its investment banking affiliates. For example, if a corporate customer of the bank were in financial distress and, in turn, the bank's loans to that firm were in jeopardy, the bank's parent would have an incentive to underwrite bonds for the firm through its securities affiliate—bonds that could then be used to pay off the troubled bank loans. Or the parent could use the securities affiliate to underwrite high-risk issues and could use the bank to extend loans to preferred customers. (Again, current law addresses these possibilities by establishing firewalls that prohibit the sharing of inside information between a bank and its affiliate.)

Other conflicts can arise from the bank's dual role as marketer of services and impartial investment advisor. Recent studies have found evidence of conflicts between promoting services and giving disinterested financial advice.³⁸ Most recently, significant conflicts between the bank's role in promoting the securities it underwrites and its role in providing disinterested financial advice have been identified and addressed. In April 2003, the nation's biggest investment firms agreed to pay a record $1.4 billion to settle charges brought by the Securities and Exchange Commission, state prosecutors, and market regulators. The firms were charged with fraud in issuing recommendations for the securities of firms they knew were in trouble, in order to acquire investment banking business. In addition to significant fines, the settlement requires the following: a clear separation of securities research from investment banking; the provision of competing, independent research for investors at no cost; better disclosure of stock rankings; and a ban on the practice of allotting initial public offering shares to favored clients.³⁹

* * *

On examination, the principal potential conflicts that are offered as a rationale for separating banking and commerce seem unlikely to pose significant risks to the safety and soundness of the bank or to the federal safety net. Firewalls—including the restrictions on the transactions between a bank and its affiliate imposed by Sections 23A and 23B of the Federal Reserve Act, the BHCA Section 106 restrictions on tying, and Federal Reserve Regulation O (which restricts transactions with insiders)—are in place to mitigate the incentives underlying the potential conflicts.⁴⁰ And market-oriented solutions—for example, competition in the markets for banking products—can also play a role in mitigating those incentives. If markets were not competitive, competition could be increased through the elimination of barriers to entry. Or if affiliation provided an informational advantage, banks without affiliates could achieve the same result, as they have done through leasing arrangements with grocery chains and other commercial firms. In short (and the point is elaborated on below), most conflict situations affecting banks can be controlled through the supervisory process and enforcement of the appropriate firewalls and need not pose excessive risk to banks or the banking system.

Concentration of Economic Power within Banking
The second kind of potential risk that is often cited as an argument for restricting the ability of banks to affiliate with nonbank firms has been the need to prevent the development of unacceptable levels of economic aggregation and power within the financial sector. When the BHCA was enacted, the concern was that the growth of unregulated bank holding companies could lead to the "undue concentration of control of banking activities."⁴¹ Since then that phrase has been used to promote two related but different goals: preventing bank monopoly power from proliferating into nonbanking businesses, and discouraging the growth of large entities.

The conventional antitrust argument for separating banking and commerce has been that banks with monopoly power will attempt to expand into nonbanking businesses to extract monopoly profits and engage in price discrimination. In practice, however, it is not likely that conglomerate integration—the combination of banks and nonbanks under a holding company—would result in monopoly rents because when markets for bank loans are competitive, it is difficult for the bank to extend market power from banking to nonbanking lines of business. Refusing to lend to the competitors of its nonbank affiliates or granting credit to its affiliates and their customers on favorable terms (as discussed above under conflicts of interest) serves only to reduce bank income. Attempts by the bank to engage in predatory pricing, by cross-subsidizing the operations of its affiliate, would work only if there were considerable barriers to entry. To the extent that banking markets are competitive, this antitrust argument for separating banking and commerce fails. Banking markets became increasingly competitive during the 1970s and 1980s; thus, the likelihood that banks would be able to extract monopoly rents was reduced. Although consolidation in banking has increased over the past decade, interstate banking and a competitive market for small banks continue to make it unlikely that monopoly power will spread from banking to nonbanking business.

The existence of limited economies of scope in banking makes it unlikely that banks and commercial firms would affiliate to the extent needed to produce an economic concentration.⁴² When Congress, in GLB, permitted combinations of large banks with large securities and insurance firms, it seemed to acknowledge that the potential for monopoly power is of less concern today and does not provide a rationale for separating banking and commerce.⁴³

The second goal mentioned above—discouraging the growth of large entities—seems equally to be based on a straw man. A fear and distrust of banks—especially large money-center banks—has long been a hallmark of U.S. political history and probably contributed to the idea that the separation of banking and commerce was necessary to prevent the growth of large conglomerates.⁴⁴ The net-public-benefits test of the BHCA instructs the Federal Reserve Board to consider, among other criteria, whether an "undue concentration of resources" would result when making its decisions regarding permissible activities for bank holding companies.⁴⁵ Over the years, the Federal Reserve Board has relied on this criteria as the basis for denying applications under Section 4(c)(8) of the BHCA when major bank holding companies have applied for approval to undertake a joint venture with a substantial enterprise.⁴⁶ Again, in light of GLB provisions mentioned above, this seems unlikely to be a justification for separating banking and commerce today.

And from the viewpoint of many community bankers, maintenance of a diversified economic system with a robust small- and middle-market business sector could be threatened by affiliations between banks and commercial firms.⁴⁷ The primary concern of these bankers is that if a large commercial entity with monopoly power, as Wal-Mart is often perceived to be, were allowed to enter banking, it would use its monopoly power to displace its banking competitors. However, the argument could also be made that a large commercial bank—not just a large commercial entity—could similarly displace its banking competitors in any given market. To the extent that there are few barriers to entry in that market, the argument that either a Wal-Mart or a large commercial bank would be able to exert monopoly power is weakened. Saunders ([1994], 239), notes that "there is no reason to expect, a priori, that the competitive behavior of the banking industry would be eroded by eliminating the commerce-banking separation. Indeed, it may be that such a policy could have a pro-competitive effect, as the number of potential entrants and potential competitors expands."

Thus, although the fear of monopoly power in banking has deep roots, it is not a sufficient reason to prohibit affiliations between banks and commercial firms.⁴⁸ Certainly concentrations of economic and political power, regardless of their source, are likely to continue to raise concerns and warrant the attention of policy makers. These concerns have traditionally been (and are best) addressed by Congress.

Safety-Net Issues
Safety-net issues arise when the bank and its affiliates (including its parent) have an opportunity or incentive to act in ways that threaten the solvency of the bank. Such a situation can exist when an insured bank enters into transactions (e.g., loans, guarantees, or other obligations or transfers) for the benefit of an affiliated person or organization and those actions endanger the safety and soundness of the bank.⁴⁹ For example, the parent organization could shift bank funds to its nonbank affiliates, or a bank could buy assets from the affiliate at inflated prices or provide a capital infusion to the affiliate through a loan at below-market rates. As a result, the parent could shift potential losses to the bank, ultimately threatening the deposit insurance funds and the taxpayer. In the context of the debate about banking and commerce, if transactions between the bank and its affiliates threaten the solvency of the bank, the fear is that the creditors of the bank's commercial affiliate(s) will be protected as a result of the federal safety net.⁵⁰

Walter (2003) expands on the circumstances under which the shifting of losses among the bank and its affiliates is likely to threaten the solvency of the insured bank and thus potentially threaten the deposit insurance funds and the taxpayer. One set of circumstances involves shifting losses among the parent's affiliates to protect the reputation of the firm; the other involves shifting losses to take advantage of limited liability.⁵¹

Under the first set of circumstances, the parent company would have an incentive to shift losses from one subsidiary to another to prevent negative information from reaching analysts and the market. For example, when the capital of the bank is greater than that of the nonbank affiliate—so that a loss shift would harm the bank but not cause it to fail—there is an incentive to shift losses to the bank, where they may go undetected and the reputation of the parent would be preserved. The reputation would be spared partly because the balance sheet of the bank might be more opaque than that of the nonbank affiliate.

Under the second set of circumstances, the shifting of losses from a more-capitalized affiliate to a less-capitalized affiliate would allow the parent to minimize its losses by taking advantage of limited liability. For example, if a nonbank affiliate incurred a loss that was larger than the capital or net worth of the bank, shifting that loss to the bank would cause the bank to fail. However, the loss to the parent would be limited to its capital investment in the failing bank. In fact, however, this shift could have a negative effect on the parent's reputation and is therefore less likely to occur than one might expect.

Because the creditors of the nonbank affiliate and the parent are more likely to exert market discipline than are the creditors of the bank, loss shifts either for reputation or for purposes of limited liability are more likely to be directed to the bank—a potential that raises safety-and-soundness concerns. The undesirable effects of shifting losses from nonbank affiliates to the bank can range from causing the bank minimal harm to causing its failure.

If the bank's creditors are aware of the potential for loss shifts, they should demand higher interest rates when they perceive a higher risk of such shifts. However, mispriced deposit insurance makes it less likely that the creditors of the bank will impose discipline by demanding higher interest rates. Thus, it is more likely that losses will be shifted to the bank.⁵² And it is precisely because these loss-shifting transactions raise safety-and-soundness concerns and potentially threaten the safety net that they have been made illegal under existing law. In particular, the firewall restrictions contained in Section 23A and Section 23B of the Federal Reserve Act require that transactions between an insured bank and its nonbank affiliates, including its parent, are on market-related, arms-length terms. Applicable to all insured depository institutions, they are intended to ensure that the loss shifting described above does not threaten the solvency of the insured bank.

Summary
The examination of the potential risks of affiliation has shown that potential conflicts of interest and the fear of monopoly power in banking do not provide sufficient justification for separating banking and commerce. Concerns about the safety and soundness of the bank and the potential expansion of the safety net exist when transactions occur in which an affiliate's financial condition is improved at the expense of the bank and those transactions threaten the solvency of the bank. The parent company and its nonbank affiliates have an incentive to shift losses to the insured bank if doing so will protect the reputation of the parent company or allow the parent to take advantage of limited liability. These incentives result when the creditors of the bank do not impose discipline on the bank.

Unchecked, these transactions and the resultant safety-net concerns would raise doubts about permitting banks to affiliate with nonbank entities, whether financial or commercial in nature. However, if regulatory discipline is imposed by the enforcement of firewalls and the prudential supervision of the insured entity, the potential harm to the deposit insurance funds and the safety net can be contained.

Managing the Risks: Firewalls and Prudential Supervision

The primary means of controlling abuse and ensuring the safety and soundness of the banking system is through the supervisory process. The goal is to balance the need for maintaining the safety and soundness of the banking system with banks' need to pursue activities and affiliations by which they can increase their profits, attract capital, and enhance their competitiveness.

The previous section mentioned several firewall restrictions: those contained in Sections 23A and 23B of the Federal Reserve Act, those contained in Section 106 of the BHCA, and the Federal Reserve Board's Regulation O. Sections 23A and 23B ensure that transactions between an insured bank and its nonbank affiliates, including its parent holding company, are on market-related, arm's-length terms. Applicable to all insured depository institutions, these restrictions are intended to ensure that the loss shifting described above does not threaten the solvency of the insured bank. Similarly, Section 106 protects from harm that may result from illegal tying. And Regulation O governs the transactions between insiders and the bank.

Other safeguards that must be in place to adequately protect the insured entity are the requirements that the bank's investment in any operating subsidiary be deducted from regulatory capital, that the bank be well capitalized following that deduction, and that the corporate separateness of the bank be protected. To achieve adequate separation, the insured entity should be financially separate—that is, it must be separately funded and have no commingled assets, and all transactions with affiliates must be at arm's length. The insured entity must also be perceived by the market to be operated separately and to be legally separate—that is, to be not responsible for the liabilities of its affiliates.⁵³

During periodic safety-and-soundness examinations, banks are examined for compliance with regulatory standards, including firewalls. As the primary federal regulator, the FDIC examines state-chartered nonmember banks; the OCC examines national banks; the Office of Thrift Supervision (OTS) examines thrift institutions; and the Federal Reserve examines state-chartered member banks. As noted earlier, the Federal Reserve also has authority to examine bank holding companies and financial holding companies. Off-site monitoring by banking regulators provides a check on the institution between examinations.

In the 1990s, regulators began to develop risk-focused supervision for those banks that are determined to be large and complex. Risk-focused supervision assesses a number of risks in each of the bank's major business lines. Because the risk profiles of large and complex banks may change rapidly, supervisors monitor risks on an ongoing basis so as to be better able to allocate supervisory resources to the areas that pose the greatest risk.⁵⁴

In combination, prudential supervision and the enforcement of regulatory standards and firewalls can provide sufficient protection for the bank and help ensure the safety and soundness of the banking system.⁵⁵ That is, these regulatory tools must be effective enough to ensure that the risk to the insurance funds is minimal, and flexible enough to allow institutions to explore the opportunities presented by affiliations with nonbank entities. The intended effect of firewalls is that the soundness of the bank not be jeopardized by an obligation to bail out an affiliate. Restrictions on the quantity and quality of transactions between the bank and its affiliates allow some synergies to be realized while minimizing the possibility that the insured bank will be harmed.⁵⁶

The result is that the existing firewalls may not be fail-safe. Firewalls are acknowledged to work well during normal times, but they are criticized for being less effective in extremis, when they may be needed most.⁵⁷ (Moreover, no firewalls have been tested since GLB made expanded affiliation possible.) And although impenetrable firewalls can be constructed, they may not be desirable. For example, as enacted in 1956, Section 6 of the BHCA achieved the complete isolation of banks within a holding company by effectively prohibiting transactions between affiliated banks, but the 1966 amendments to the BHCA repealed the prohibition.⁵⁸

Two Regulatory Models

How then should affiliations be handled? Although separating banking from commerce would prevent the risks described above, it would do so in a heavy-handed way and would prevent the economically preferred market-based outcome from being realized. As Walter (2003) noted, "Maintaining a wall separating banking and commerce at best addresses a symptom of an uncompetitive market rather than the lack of competition itself."⁵⁹

The alternative to prohibiting affiliations between banking and commercial firms is to regulate the affiliations so that potential harm to the safety net is contained. There are two regulatory models or long-term strategies for accomplishing the objective. In this context, the two models are distinguished by the extent to which the entire enterprise, including the parent of the insured entity, is subject to federal oversight. The first model reflects the view that federal oversight of the banking organization from the top down—that is, from the parent down to its subsidiaries, both bank and nonbank—is necessary if the safety net is to be protected. The second model reflects the view that regulatory scrutiny can be accomplished from the bank or insured entity up: adequate safeguards will make it possible to protect the insured entity from the risks and conflicts that arise from affiliations without the need for explicit oversight of the parent.

The Top-Down Approach to Regulating the Banking Organization
The top-down model—advocated by the Federal Reserve Board (among others), codified by the BHCA, and amended most recently by GLB—features federal oversight in the form of consolidated supervision of a bank holding company and umbrella oversight of a financial holding company.⁶⁰ As mentioned, the rationale for supervision of financial holding companies is to ensure that the safety net is protected from the risks associated with affiliations outside of banking. The logic of this model suggests that, as the market creates pressure for additional commercial activity to be associated with banking, there is a need to ensure that those activities will be subject to federal oversight.

Proponents of the top-down model argue that it maintains a separation of banking from commerce by limiting the ability of banks and banking organizations to own, or be owned by, nonfinancial or commercial firms. This limitation is achieved in three ways. First, affiliations among financial services providers are permitted only under the organizational structure of the financial holding company, which is subject to federal oversight (including functional regulation of certain nonbank affiliates and umbrella supervision of the entire organization by the Federal Reserve). Second, the Federal Reserve (in conjunction with the Secretary of the Treasury), has effective control over the determination of permissible activities for financial holding companies—thus, over banking and commerce.⁶¹ Finally, by eliminating the unitary thrift holding company charter, GLB precludes further commercial ownership of thrifts. But questions have been raised as to whether GLB maintains or undermines the separation of banking and commerce (see below).

Oversight of the Banking Organization. Traditionally, those who argue that affiliations between banking and commerce should be prohibited believe that the reliance on firewalls and prudential supervision alone is not sufficient to protect the insured entity.⁶² These advocates of separation question whether firewalls can be strong enough to prevent unacceptable levels of risk from harming the insured entity, yet flexible enough to permit the economic advantages of affiliation to be realized. They also question the ability of firewalls to ensure the corporate separateness of the insured entity.⁶³ Accordingly, supervision of the insured entity and enforcement of firewalls to protect it from the risks posed by affiliation must be supplemented by consolidated or umbrella supervision of the entire banking organization.⁶⁴

As noted above, bank holding companies became subject to consolidated supervision by the Federal Reserve under the BHCA. Under consolidated supervision, separate units of a holding company are supervised as one entity. The consolidated supervisor has direct oversight of the parent and its subsidiaries so that the relationship between nonbank affiliates and the insured entity can be controlled. Under GLB, the Federal Reserve received the power to be the umbrella supervisor of the financial holding company and in that capacity has various authorities.⁶⁵ But for the purposes of umbrella supervision, functionally regulated nonbank affiliates are not treated as banks, and the Federal Reserve is directed to rely primarily on the information provided by the nonbank affiliates' functional regulators.⁶⁶

Consolidated supervision has been criticized on several counts. Some of the critics have noted that it can be viewed as a "vote of no confidence in firewalls."⁶⁷ If firewalls can effectively protect the insured entity, consolidated supervision is unnecessary. The argument is also made that consolidated supervision signals the market that regulators expect affiliates to be managed as integrated entities. As a result, ensuring effective separation of the insured entity from the risks posed by its affiliates may become harder as supervisors are more likely to focus on the integrated entity rather than the insured depository institution. Moreover (the argument goes), the requirements of consolidated supervision reduce the flexibility of bank or financial holding companies to adapt to a rapidly changing financial environment and to best meet the needs of their customers.⁶⁸ The view that consolidated supervision need not extend to the owners of banks was clearly articulated by the interagency study issued in 1991 by the U.S. Treasury, which made broad recommendations for modernizing the financial services industry.⁶⁹ The Treasury study noted that beyond certain necessary safeguards designed to ensure that the safety net was not extended beyond the bank, cumbersome bank-like regulation was not necessary for the financial services holding company that was then being proposed.⁷⁰ The study argued that "it is practically infeasible for a bank supervisor to effectively regulate a complex and diverse range of businesses. Bank regulation should be concentrated on the bank, not on protecting a diversified [financial services holding company] that should be subject to normal market discipline."⁷¹ The study noted that consolidated regulation of the holding company ran the risk of being viewed as implicit government backing of holding companies, increasing the taxpayer's exposure and potentially expanding the safety net beyond the insured entity. The study also noted that full holding company regulation would deter nonbanking firms from investing in banks because potential investors would deem too high the price of having bank supervisors regulate all nonbanking activities. Moreover, the study noted that none of the hypothetical problems of combining banking and commerce was evident among commercial companies that owned depository institutions at that time, notably unitary thrifts, nonbank banks, and industrial loan companies.⁷²

Similarly, the GLB requirement that financial firms submit to umbrella oversight through the financial holding company structure may deter these firms' entry into banking. Although GLB's functional regulation provisions appear initially to shield nonbank firms, the Federal Reserve's authority and powers remain extensive, and from the viewpoint of these firms, GLB has restricted the incentives of the marketplace.⁷³ A further deterrent may be the Federal Reserve's source-of-strength doctrine, which requires a holding company to provide financial assistance to its troubled subsidiary banks. Although the doctrine was restricted by the Fifth Circuit Court of Appeals in a case in which the Federal Reserve ordered a holding company to inject capital into a failing bank, under current regulation the source-of-strength doctrine remains the Federal Reserve's policy.⁷⁴ By making investments in bank equities less attractive, the policy could have the effect of raising the organization's cost of capital. And because the policy is directed primarily at the corporate owners of banks, it would lead to the differential treatment of individual owners, for presumably they would not be held to the standard.⁷⁵

Determination of Permissible Activities. Under GLB, the Federal Reserve plays the dominant role in determining permissible activities for the financial holding company and its subsidiaries. Beyond an initial set of permissible activities, GLB authorizes the Federal Reserve Board, in conjunction with the Secretary of the Treasury, to determine additional permissible activities—those that are financial in nature or incidental to such financial activities. The Federal Reserve alone is authorized to determine the set of activities that are complementary to financial activities, as long as they do not pose a substantial risk to the safety and soundness of the insured entity or the financial system.

The dynamics of expanding the list of permissible activities are different under GLB than under the BHCA. Under GLB the test is "financial in nature or incidental to," and unlike the "closely related to" standard of the BHCA, there is no net-public-benefits test. Once defined as permissible, an activity is open to financial holding companies and financial subsidiaries with only a post-entry notification to the Federal Reserve required. As a result, subsequent competitive evaluations are not possible. And, as noted earlier, GLB also directs the Federal Reserve to permit an unspecified set of commercial activities defined as complementary, and permits financial holding companies to engage in merchant banking.⁷⁶

Through the provisions for determining what activities are financial, incidental, or complementary—and are thus permissible for financial holding companies—GLB has shifted the line of separation: instead of drawing it between banking and commerce, GLB has drawn it between finance and commerce. On this basis, some argue that GLB effectively undermined the policy rationale for separating banking and commerce.⁷⁷ Moreover, the process of defining what is financial yet not commercial considerably weakens the congressional intent to maintain a separation of banking and commerce. The line becomes adjustable in response to changes in markets and technology, as determined primarily by the Federal Reserve. The result is a blurring of the distinctions between banking, finance and commerce, and without operational limits to expansion, GLB suggests a slow but accelerating integration of banking and commerce.⁷⁸

But it is hard to argue that the potential risks posed by affiliations between securities firms and banks are less risky than those posed by affiliations between banks and other nonbank or commercial entities. For example, with regard to the use of credit or the benefits to be gained from affiliation with a bank, the activities of a securities firm do not differ significantly from those of a commercial firm.⁷⁹ If it is imperative to keep banking separate from commerce, it should be no less important to separate banking from securities activities. By permitting affiliations among banking, securities, and insurance, GLB tacitly acknowledges that the safety-and-soundness risks posed by these affiliations are manageable. This acknowledgment makes it hard to defend the principle of separation.

Elimination of the Unitary Thrift Holding Company Charter. GLB placed new restrictions on the ability of banking and commerce to mix. Specifically, it ended the ability of commercial firms to own a single thrift in a unitary thrift holding company, although existing holding companies were grandfathered.⁸⁰

Supervised by the OTS, unitary thrift holding companies are subject to prudential supervision and firewall restrictions⁸¹ and have long operated without raising safety-and-soundness concerns. For example, thrifts that were part of diversified holding companies were not significant sources of losses during the savings and loan crisis of the 1980s.⁸² Nor have they been recipients of a significant number of enforcement actions.⁸³ In addition, thrifts in unitary thrift holding companies have tended to outperform other thrifts because of the greater diversification of their revenue streams, loan and asset portfolios, and funding sources. The mixing of banking and commerce conducted in the unitary thrift holding company has not been shown to pose undue risk to the safety net.⁸⁴ Despite this, those opposed to the mixing of banking and commerce consider it unsuitable for a diversified holding company to own a single thrift.⁸⁵

The enactment of GLB restrictions on diversified holding company ownership of thrifts closed a long-standing avenue through which banking and commerce have successfully mixed. Like ILCs, unitary thrift holding companies operated outside of the bank holding company structure and outside of supervision by the Federal Reserve. When GLB eliminated this corporate structure, it narrowed the options available for mixing banking and commerce.

The Bank-Up Approach to Regulation: Protection of the Insured Entity
Sometimes referred to as a bank-up or safeguard approach to supervision, the other model for containing potential risks focuses on protecting the insured entity—and, in turn, the insurance funds—at the bank level rather than at the parent's level. Properly enforced, firewalls and regulatory safeguards can serve to ensure legal and financial separation and to promote market separation. These protections can be confined to the insured entity, and regulatory and supervisory oversight of the parent and nonbank affiliates becomes unnecessary. Because the insured entity can be adequately protected by safeguards applied at the bank level, this model will be called the bank-up model.⁸⁶

By focusing on the bank itself rather than attempting to oversee the entire banking organization, bank supervisors should be able to defend adequately against any tendency by the owners of the bank to aid their nonbank affiliates. To do this, the banking supervisor requires certain powers, including the authority to monitor compliance on both sides of all transactions between the insured entity and its affiliates (including its parent), the authority to require that the insured entity and its affiliates report such transactions, and—in the case of nonbank affiliates—the authority to require that financial statements or other relevant information be made available to the primary bank regulator.⁸⁷

As the primary regulator of state-chartered, nonmember banks and as the deposit insurer, the FDIC has various powers that allow it to ensure the safety and soundness of the insured entity and, by extension, to ensure the safety of the deposit insurance funds. When the FDIC is the primary federal regulator—for example, for ILCs—the necessary protections can be provided without consolidated oversight of the insured entity's owners. Among the powers the FDIC has are the authority to examine both sides of transactions between the bank and its affiliates and to examine the bank and any affiliate, including the parent, as may be necessary to determine not only the relationship between the insured entity and the affiliate but also the effect of such relationship on the insured entity. When the parent is subject to the reporting requirements of another regulatory body (e.g., the Securities and Exchange Commission or a state insurance commissioner), the FDIC has agreements in place to share information with that regulator.⁸⁸ Moreover, in examining any insured depository institution, the FDIC has the authority to examine any affiliate of the insured entity, including its parent company, as may be necessary to determine the relationship between the insured entity and the affiliate, and the effect of the relationship on both of them.⁸⁹

The regulation and supervision of ILCs illustrate how the bank-up model can effectively protect the insured entity and the insurance funds without subjecting the entire organization to consolidated or umbrella supervision.⁹⁰ The state of Utah, as home to approximately one-half of all ILCs and more than 80 percent of industry assets, provides a case in point.

ILCs became eligible for federal deposit insurance under the Garn-St Germain Act of 1982. In 1987, the Competitive Equality Banking Act expressly exempted ILCs from the BHCA and from oversight of the parent organization by the Federal Reserve.⁹¹ GLB did not repeal this exemption. Generally ILCs are authorized to engage in traditional financial activities that are available to all charter types. They may make all kinds of consumer and commercial loans and may accept federally insured deposits, but not demand deposits if total assets exceed $100 million. They may be original issuers of Visa or MasterCard credit and debit cards and may fund their operations with Federal Home Loan Bank borrowings. If an ILC is organized as a limited-purpose or credit-card institution, its products and services are limited to those specified by its charter.

ILCs are subject to the same regulatory and supervisory oversight (including the laws and regulations pertaining to bank safety and soundness and consumer protection) as chartered banks. They are subject to the same or higher capital requirements and the same regulations affecting transactions between the insured entity (the ILC) and its parent and affiliates. However, because of the ILC exemption from the BHCA, the activities and powers permitted under the ILC charter are restricted less than those under other charters. Given its flexibility, the ILC charter has been an attractive choice for companies that are not permitted to, or choose not to, become subject to the restrictions of the BHCA. As a result, it is not surprising that the parent companies of ILCs include a diverse group of financial and, where permitted, commercial firms.

The example of the ILCs offers one answer to the question of whether a bank regulated at the bank level can be insulated and isolated from the parent company's improprieties. G. Edward Leary, commissioner of the Utah Department of Financial Institutions, is among those who have argued that it can, noting that the contrary case has not been made. Leary argues that the collaboration between Utah and the FDIC has resulted in the effective regulation and supervision of the state's ILCs and serves as a working example of how well the bank-up approach can work.⁹²

By Leary's admission, the supervision of ILCs is an evolving regulatory dynamic in the sense that the regulations must evolve to meet the changes in the industry. This responsive evolution is most visible in the approval orders for de novo ILCs, which contain many of the prudential safeguards under which the ILC will operate for the life of its charter. Among the safeguards for Utah's ILCs are requirements designed to maintain the autonomy of the board of directors and management and their independence from the parent company. To achieve autonomy and independence, the ILC's management must act in the best interest of the ILC; must maintain accurate and reliable accounting records on-site—records on which to base its decisions; must retain authority to set policy and make decisions for credit underwriting; must ensure that all transactions with the affiliate or parent corporation are at arm's length; and must have sufficient personnel and resources to carry out its decisions.⁹³ The state of Utah requires that the parent company register with the Utah Department of Financial Institutions and be subject to the department's jurisdiction and examination authority. As noted above, even though the FDIC is not designated the umbrella supervisor of ILCs, it does have the authority to examine the parent company of the ILC to determine the relationship between the parent and the ILC and the effect of that relationship on the ILC. That is, it has the authority to "supervise the organization" from the bank up.

Because ILCs are ably and effectively regulated and supervised from the bank up—both at state and federal levels—it is argued that they pose no greater safety-and-soundness risk than other charter types. Troubles in ILCs have not stemmed from issues pertaining to permissible activities or commercial affiliations or from the regulatory structure under which they operate, but from faulty strategic or tactical decisions.⁹⁴ In short, the ILC charter does not represent an inappropriate mixing of banking and commerce. It is important that the ILC charter should not be seen as a loophole, but as a viable charter type and supervisory option.

The Example of Conseco. The bankruptcy of the corporate owner of an ILC—Conseco Inc.—but not of the ILC itself illustrates how the bank-up approach can effectively protect the insured entity without there being a BHC-like regulation of the parent organization. Conseco Inc. was originally incorporated in 1979 as Security National of Indiana Corp. After several years of raising capital, it began selling insurance in 1982. Security National of Indiana changed its name to Conseco Inc. in 1984, after its 1983 merger with Consolidated National Life Insurance Company. Conseco Inc. expanded its operations throughout the 1980s and 1990s by acquiring other insurance operations in the life, health, and property and casualty areas.⁹⁵ Conseco Inc. was primarily an insurance company until its 1998 acquisition of Green Tree Financial Services. A diversified financial company, Green Tree Financial Services was one of the largest manufactured-housing lenders in the United States.⁹⁶ Upon acquisition, it was renamed Conseco Finance Corporation. Included in the acquisition were two insured depository charters held by Green Tree Financial Services—a small credit-card bank chartered in South Dakota⁹⁷ and an ILC chartered in Utah. Both of these institutions were primarily involved in issuing and servicing private-label credit cards, although the ILC also made some home improvement loans. The ILC—Green Tree Capital Bank—was chartered in 1997 and changed its name to Conseco Bank in 1998 after the acquisition. Conseco Bank was operated profitably in every year except the year of its inception, and grew its equity capital from its initial $10 million in 1997 to just over $300 million in 2003. Over the same period, its assets ballooned from $10 million to $3 billion.

Conseco Bank was supervised by both the Utah Department of Financial Institutions and the FDIC. Despite the financial troubles of its parent and the parent's subsequent bankruptcy (filed on December 18, 2002), Conseco Bank's corporate firewalls and the regulatory supervision provided by Utah and the FDIC proved adequate in ensuring the bank's safety and soundness. In fact, $323 million of the $1.04 billion dollars received in the bankruptcy sale of Conseco Finance was in payment for the insured ILC—Conseco Bank, renamed Mill Creek Bank—which was purchased by GE Capital.⁹⁸ As a testament to the Conseco Bank's financial health at the time of sale, the $323 million was equal to the book value of the bank at year-end 2002.⁹⁹ Thus, the case of Conseco serves as an example of the ability of the bank-up approach to ensure that the safety and soundness of the bank is preserved.

Policy Implications

This section discusses whether umbrella oversight of the parent company is necessary and explores how the debate about banking and commerce is affected by the increased complexity of banking organizations and how important it is for the banking system to have regulatory alternatives.

A Need for Umbrella Oversight of the Parent?
If the risks posed by mixing banking and commerce can be contained through adequate safeguards, affiliations need not be prohibited. But is it possible to permit affiliations between banking and commerce without imposing federal regulation or oversight on the bank's parent company and affiliates? In other words, under what conditions is umbrella oversight of the parent necessary?

The purpose of umbrella oversight is to protect the insured entity from the risks posed by its affiliates. Over the years, varying opinions on the need for umbrella oversight have been expressed. From the perspective of the Federal Reserve, umbrella oversight provides the necessary framework that allows the risks associated with an organization's consolidated activities to be monitored and restrained. Umbrella oversight protects the insured entities in the organization and, in turn, the safety net. It also makes financial crises and risk to the financial system easier to prevent.¹⁰⁰

Walter (2003) reached the conclusion that mixing banking and commerce need not be prohibited but argues that umbrella oversight of the entire organization is necessary. He notes that umbrella oversight is intended to mimic the types of limitations that private creditors would impose on risky affiliations.¹⁰¹ When uninsured creditors are aware of increased riskiness on the part of their debtor, they demand compensation for the added risk. In the face of increased risk taking by a nonbank affiliate, supervisors would similarly impose restrictions or other penalties to compensate for the added risk posed by the affiliate. For example, the supervisor could reduce the bank's supervisory rating or impose restrictions on transactions with the bank's affiliates. In doing so, the supervisor would be mimicking the monitoring behavior of the private creditor in the absence of deposit insurance and a federal safety net. In Walter's model, supervision of the entire banking organization is performed by the Federal Reserve in a top-down model.

Others have expressed the view that umbrella supervision of the entire organization is neither necessary nor warranted.¹⁰² If commercial firms that choose to enter the banking business were subjected to umbrella supervision, growing amounts of economic activity would be brought into a regulatory framework that was designed to administer the nation's financial safety net. Instead, federal oversight could be focused on containing the risks posed by such mixing of banking and commerce—the risk that losses would be shifted to the insured entity and, in turn, to the deposit insurer—through the use of firewalls and prudential supervision directed at the insured entity. Confining regulatory oversight to the bank can achieve effective regulation of the insured entity without unwarranted regulatory intrusion into the marketplace.

As noted above, experience has shown that the bank-up model, with proper safeguards, can work. In the bank-up regulatory model, supervision is performed by the primary regulator, which stands in the place or performs the role of the uninsured creditor. Importantly, regulatory discipline can be exerted to protect the safety net without the parent organization and the bank's nonbank affiliates being subjected to federal supervisory oversight. Under these conditions, regardless of holding company affiliation or size, banks should be entitled to choose the corporate structure that best suits their business needs.

Separation and the Increased Complexity of Banking Organizations
A significant change during the past two decades has been an increase in both the size and the degree of complexity of financial organizations. Some observers have noted that on a global scale, in the absence of adequate controls, combinations of banking and finance can produce large losses that must be borne by society at large.¹⁰³

How does the greater complexity of financial organizations affect the debate about banking and commerce? Given the increased complexity, what (if anything) is needed to ensure that the risks posed by the mixing of banking and commerce can be contained within a framework of regulatory choice? Are current regulations sufficient? Or can they be improved?

Before GLB was passed, nonbank affiliates were generally quite a bit smaller and less complex than their bank affiliate. As a result, the incentives for the parent to shift losses to the bank affiliate in order to take advantage of limited liability were lessened. Today, under a financial holding company, banks are able to affiliate with securities and insurance firms that are likely to be as large as, if not larger than, the bank itself. The result is often an organization that is both large and complex, and is likely operated as an integrated entity that manages risk across business lines, rather than within legal entities. Thus, the likelihood may be greater that these large and complex financial organizations may attempt to shift losses to the bank and the insurance funds for limited-liability purposes.¹⁰⁴

As financial organizations increasingly rely on risk-control strategies that view the activities of the organization in toto, it is all the more important that regulators have the ability to assess the risk posed to the insured entity.¹⁰⁵ The framework of umbrella oversight in intended to provide the necessary transparency regarding the activities, practices, and inter-company dealings that affect the distribution of risk across the financial holding company, and to serve as a supplement to supervision by legal entity in that case. However, some have argued that the necessary degree of oversight or monitoring might be better determined by the degree of complexity within the organization, rather than solely on the basis of affiliation.¹⁰⁶ For example, organizations that combine banking and finance, where business lines are managed at the holding company level, might warrant more of such oversight than organizations that combine banking and commerce, where the insured entity is clearly separate.

Outside the financial holding company structure, however, the transparency necessary to properly assess risk and protect the insured entity should be accomplished without requiring the organization (parent and affiliates) to be subject to a top-down form of umbrella supervision. Rather, any monitoring should be the responsibility of the functional regulator, and policy makers should consider whether additional powers may be required to ensure sufficient transparency.¹⁰⁷ A clear benefit of this approach is that the nonbank economic activity associated with the mixing of banking and commerce would continue to be driven by the market rather than by regulation.

The Importance of Regulatory Alternatives
Throughout U.S. banking history, the states have chartered, regulated, and supervised banks. The existence of state-chartered depository institutions, including ILCs, remains integral to the checks and balances of the dual banking system. In meeting the banking needs of their communities, state-chartered institutions have fostered creativity and experimentation. Part of the reason the state charter remains a viable banking option is that it allows for innovation in a locally controlled environment that has traditionally limited systemic risk. This capacity for innovation is particularly true of the ILC charter.

It is important that there be more than one approach to the regulation and supervision of banks. A key attribute of the dual banking system is the insured entity's ability to choose the supervisory structure under which it operates: the ability to choose contributes to a competition in excellence among bank regulators. Through its role as primary regulator of state-chartered nonmember banks and ILCs, the FDIC provides the bank-up regulatory alternative for organizations and individuals that choose not to be regulated by the Federal Reserve under a holding company structure. Thus, this model offers greater flexibility for corporate enterprise, while managing the risks posed by a mixing of banking and commerce. Without this alternative regulatory structure, the ability of the market to meet the demands of consumers could be severely restricted.

Conclusions

Does the mixing of banking and commerce constitute good public policy? The evidence suggests that the answer is a qualified yes: with adequate safeguards in place, the careful mixing of banking and commerce can yield benefits without excessive risk. Because the main potential risk of mixing banking and commerce—the shifting of losses that may threaten the insured institution and the safety net—can be contained through the use of adequate safeguards and firewalls, the separation of banking and commerce does not appear to be justified.

Furthermore, there is no evidence of a long-term separation of banking from commerce in U.S. banking history. Certainly the activities permitted to banks have always been subject to prohibition, but affiliations with nonbank firms have not been prohibited until relatively recently. Nevertheless, the current prohibitions on corporate ownership of banks have been justified on the grounds that banking and commerce have always been separate.

Despite the current prohibitions, the current regulatory structure makes it possible for banking and commerce to continue mixing in many ways. In addition, the current trend in the market place is toward more combinations of banking and commerce. As FDIC Chairman Powell recently noted, "This trend is nothing more than the natural outgrowth of dynamics that have been under way in banking and bank regulation over the last two decades."¹⁰⁸ The issue facing policy makers is how these combinations of banking and commerce will be regulated. Specifically, will increasing amounts of commercial activity be subject to umbrella supervision, or will the insured entity be the focus of supervision?

Is umbrella oversight of the entire organization necessary? The evidence suggests that the answer is a qualified no. Given the important role that regulatory choice has played and continues to play in the U.S. banking industry, there should be no need either to reconcile or to choose between the two approaches to regulating banking organizations. Each approach performs its role within the current regulatory system. Given the increased complexity of many banking and financial organizations, it is important that the risks to the insurance funds and the safety net are understood and that firewalls remain effective. In particular, there may be a role for added information sharing and disclosure within the current regulatory structure. Importantly, that added oversight could be performed under either the top-down or the bank-up regulatory model. Regulators and policy makers should consider what additional powers, if any, are needed to be able to effectively ensure corporate separateness of the insured entity, while also ensuring regulatory choice about how the banking enterprise is regulated.

* * *

^*The author is a senior financial economist in the Division of Insurance and Research at the Federal Deposit Insurance Corporation. The author would like to thank Christine Bradley, Valentine (Missy) Craig, and Rose Kushmeider for insightful comments and careful review of earlier drafts, and Steven McGinnis for research assistance. Thanks also go to those individuals from the FDIC's Division of Supervision and Consumer Protection who provided valuable comments.

¹ ILCs are discussed more fully in a later section of this paper.

² In April 2003, the House passed both H.R. 758 and H.R. 1375. The former is the proposed Business Checking Freedom Act, which would allow banks to pay interest on business demand deposits and would permit ILCs to offer their corporate customers interest-bearing negotiable order of withdrawal (NOW) accounts. The latter is the proposed Financial Services Regulatory Relief Act of 2003, which would remove the remaining regulatory barriers to interstate de novo banking: banks and ILCs would be allowed to use start-up branches to cross state lines. In March 2004, the House amended H.R. 1375 to restrict the ability of certain ILCs to branch interstate: only ILCs that had been established before October 1, 2003, and were owned by companies such that no more than 15 percent of income is derived from nonfinancial sources would be permitted to branch interstate. The amendment effectively prevents commercial firms such as Wal-Mart from using the ILC charter to develop a branch banking business. As of March 2004, none of these issues had been addressed by the Senate.

³ See, for example, Greenspan (2003).

⁴ See Powell (2003a).

⁵ See Powell (2003b).

⁶ On July 16, 2003, the FDIC held a symposium at the National Press Club in Washington, D.C., entitled "The Future of Banking: The Structure and Role of Commercial Affiliations," where the issue of affiliation between banks and commercial firms was discussed. Several of the papers presented there are referenced in this paper.

⁷ Such oversight of bank holding companies has been the purview of the Federal Reserve Board under the Bank Holding Company Act (BHCA).

⁸ See, for example, Corrigan (1987, 1991) and Jorde (2003).

⁹ See, for example, FDIC (1987a, 1987b) and Wallison (2003).

¹⁰ See FDIC (1987b), 3.

¹¹ Ibid.

¹² See Golembe (1997) for an overview of the policy issues. See also Blair (1994, 2004), FDIC (1987a), Halpert (1988), Hammond (1936, 1957), Haubrich and Santos (2003), Redlich (1951a), Shull (1999), and Symons (1983).

¹³ Limited-purpose consumer banks, which are national or state-chartered banks and operate under certain restrictions, are discussed in Yom (2004). Industrial loan companies are discussed below under "The Bank-Up Approach to Regulation" and in West (2004).

¹⁴ Because most states in the early part of the twentieth century prohibited branch banking, chain banking provided a way for one or more individuals to hold control in a chain of several banks. When permitted by law, chain banking organizations often were turned into branch banks or evolved into bank holding companies. See Klebaner (1990).

¹⁵ The European-style universal bank has greater freedom to be owned by, and to own, nonfinancial firms (Barth, Caprio and Levine [2000]). Barth, Caprio and Nolle (2004) present a relative ranking of countries by permissible banking activities and ownership restrictions, noting that countries generally place greater restrictions on the ability of banks to own nonfinancial firms, than on the ability of nonfinancial firms to own banks. The EU countries are ranked among the least restrictive countries, while the United States ranks as one of the most restrictive, although the data used was compiled before the passage of GLB. Shull and White (1998), 14, note that a pure universal bank requires neither subsidiaries, affiliates, nor holding companies. Traditional banking and nontraditional activities are carried out in an integrated manner; that is, they are not separated by firewalls nor the legal doctrine of corporate separateness. Universal banks typically provide both short-term banking credit and underwriting and equity investments for intermediate and long-term capital formation. They are characterized by close and long-term relationships between the bank and its commercial and industrial customers. The Japanese keiretsus are conglomerate groupings in which banks are linked to their client companies through equity ownership. Concerns that the mixing of banking and commerce would produce a concentration of power have evoked comparisons with these systems. The comparisons, however, are misleading, for the close ties among the government, commercial firms, and banks found in the Japanese keiretsus and between European universal banks and commercial firms are unlikely to be replicated in the United States. U.S. capital markets developed early and have been an important source for corporate funding, especially relative to European markets. And U.S. banking law prohibits banks and commercial firms from being both creditors and shareholders. As an example, if Citigroup were acquired by General Electric (or vice versa), the bank subsidiary (or affiliate) would continue to be prohibited from owning stock in the other.

¹⁶ The FDIC was among those calling for such financial modernization; see FDIC (1987a). See Corrigan (1987) for the perspective of the Federal Reserve. The U.S. Treasury also made recommendations for modernizing the financial system; see U.S. Department of the Treasury (1991).

¹⁷ Before passage of GLB, banking organizations had been permitted to engage in the following: the post—Glass-Steagall securities activities in which bank holding companies were permitted to engage through Section 20 subsidiaries, the securities activities that the states permitted to subsidiaries of state-chartered nonmember banks, and activities that the OCC permitted to operating subsidiaries of national banks under its interpretation of the National Bank Act. Also included were the insurance activities of state-chartered banks and the insurance agency activities permitted to banking organizations and national banks operating in communities with populations under 5,000.

¹⁸ GLB provided for affiliations between qualifying bank holding companies—called financial holding companies—and securities and insurance firms. The Federal Reserve Board was designated the umbrella regulator of the financial holding company, while the bank, securities, and insurance affiliates were to be supervised on a functional basis. The Federal Reserve (in conjunction with the Secretary of the Treasury) was authorized to determine the set of activities—those that are financial in nature, or incidental to such financial activities—that are permitted within the financial holding company. In addition, the Federal Reserve can determine that complementary activities that do not pose a significant risk to the safety and soundness of depository institutions or of the financial system generally are permissible activities. GLB also recognized merchant banking as financial in nature and a permissible activity for financial holding companies.

¹⁹ See, for example, Haubrich and Santos (2003).

²⁰ Haubrich and Santos (2003), 112.

²¹ Halpert (1988), Saunders (1994), Shull and White (1998), and Krainer (2000), among others, discuss the potential benefits and costs of mixing banking and commerce. (Potential risks are discussed in the next section of this paper.)

²² See Saunders (1994), Shull (1999), and Krainer (2000) for a discussion of the literature.

²³ See Krainer (2000).

²⁴ Shull (1999) and Saunders (1994) discuss these incentives. Shull notes that many of the same issues that arose in the debates over bank expansion into finance (the securities and insurance businesses) are relevant to the debate about banking and commerce.

²⁵ In 1991, the U.S. Department of the Treasury published an interagency study that made recommendations for modernizing the financial system (see U.S. Department of the Treasury [1991], 54-61). The study recognized the benefits of lowering the barriers between banking and commerce, but it did not recommend lowering them evenly. Affiliations between banking and commercial firms were recommended partly as a way to infuse capital into a then-weak banking system. The study recommended that commercial firms be allowed to own banks indirectly through a financial services holding company, although banks and bank holding companies were not to be permitted to acquire commercial firms as subsidiaries or hold equity claims on commercial firms on their balance sheets. Banks and financial firms would have been able to affiliate with each other.

²⁶ One example is DaimlerChrysler, which was formed in 1998 from the merger of Germany's biggest industrial group, Daimler-Benz, with Chrysler Corporation. The mixing of banking and commerce in this case arises from the ownership position held by Germany's Deutsche Bank in the former Daimler-Benz. DaimlerChrysler's long-range plans called for the creation of a global entity that would include automobile leasing and finance, information technology, real estate, and telecommunications, into one financial-services provider.

²⁷ Haubrich and Santos (2003), 159, note that although the merchant-banking provisions do not allow banks to own or operate nonfinancial firms, the possibility of bank control of nonfinancial firms remains.

²⁸ Other literature suggests that there are limits to the synergies between commercial and investment banking. See Craig (2004).

²⁹ FDIC (1987a), 46. Conflicts of interest are also discussed in Halpert (1988) and Walter (2003).

³⁰ Craig (2004) notes that the recent expansion of commercial banking into investment banking, as allowed under GLB, has increased the potential for conflicts of interest.

³¹ Halpert (1988), 508-9, notes that concern over this potential conflict of interest was expressed during the congressional debates that preceded the 1970 amendments to the BHCA, which expanded the BHCA's purview to all bank holding companies. At that time, one-bank holding companies were free to engage in commercial activities through nonbank affiliates without being subject to the BHCA. Supporters of the amendments claimed that deposit and asset growth of one-bank holding companies threatened the availability of credit to nonaffiliates, although no evidence of such behavior was presented.

³² Federal Reserve Act, ch. 6, §§23A and 23B (1913) (codified as amended at 12 U.S.C. §§371c and 371c-1 (2001)).

³³ Bank Holding Co. Act Amendments of 1970, Pub. L. 91-607, Title I, §106(a)-(h), 84 Stat. 1766 (1970) (codified as amended at 12 U.S.C. §§1971-78 (2001)).

³⁴ Legal tying might be better described as cross-selling.

³⁵ Craig (2004) notes that the Federal Reserve Board recently clarified the following as traditional bank products: loans, discounts, deposits, trust services, cash management, custodial services, payroll services, settlement and wire transfer services, and discretionary asset management.

³⁶ Several banks have recently faced losses from lines of credit that were extended to corporate customers in return for receiving that corporation's underwriting business. In this sense, legal tying or cross-selling can lead to losses that could threaten the bank's safety and soundness. See Craig (2004).

³⁷ See Board of Governors of the Federal Reserve System (2003), U.S. General Accounting Office (2003), and Office of the Comptroller of the Currency (2003).

³⁸ Craig (2004) reviews these studies. 39 Although a settlement was reached, it should be noted that the firms did not concede that they were guilty of the charges. Craig (2004).

⁴⁰ The statutory basis for Regulation O (12 CFR 215) is Federal Reserve Act, §§22(g) and 22(h) (1913) (codified as amended at 12 U.S.C. §§375a and 375b).

⁴¹ Halpert (1988), 500ff.

⁴² Halpert (1988), 507.

⁴³ Wallison (2000, 2003) argues that GLB, in effect, says that none of the reasons advanced against commercial ownership of banks are valid.

⁴⁴ Walter (2003), 13-15, notes that the legislative history of the BHCA indicates that the Congress' intent was to guard against the undue concentration of control in banking activities and that this has been interpreted as a concern over the proliferation of conglomerate monopoly, where both banks and nonbanks are combined under one holding company.

⁴⁵ The BHCA establishes a net-public-benefit test under which the Federal Reserve Board must consider "whether the performance of the activity by a bank holding company or a subsidiary of such company can reasonably be expected to produce benefits to the public, such a greater convenience, increased competition, or gains in efficiency, that outweigh possible adverse effects, such as undue concentration of resources, decreased or unfair competition, conflicts of interest, or unsound banking practices." 12 U.S.C. 1843(j)(2)(A).

⁴⁶ For examples of such decisions, see Halpert (1988), 506, and Walter (2004), 14.

⁴⁷ See Jorde (2003) on the viewpoint of community bankers. She notes, for example, that permitting banking and commerce to mix would run counter to the U.S. ideals of separation and dispersion of political and economic power and would exacerbate the current trends of consolidation in banking and other industries.

⁴⁸ U.S. Department of the Treasury (1991), 57, noted that the allocation of credit and the concentration of economic power were best addressed by means other than prohibiting the mixing of banking and commerce.

⁴⁹ Transactions with affiliates—commercial or otherwise—need not pose safety-net concerns. For example, upstreaming dividends to the bank's parent organization would be acceptable, provided the dividends were reasonably related to the bank's existing capital and earnings potential. However, when transactions benefit a related party and are detrimental to the viability of the insured bank, the safety net can be threatened. See FDIC (1987a), 87.

⁵⁰ Conversely, the parent could engage in activities that benefited the bank at the expense of its affiliates. It is argued generally that this conflict is of less concern because fewer safety-and-soundness issues surround most nonbanking firms. When the bank is allowed to affiliate with other businesses or to own nonbank subsidiaries, that affiliate or subsidiary can be sold to generate a source of added capital for the bank. See, for example, FDIC (1987a), chap. 5, "Conflicts of Interest." See also Jones and Kolatch (1999) for a discussion of the relative benefits of the bank subsidiary model.

⁵¹ Walter (2003) discusses a third circumstance that has been cited as justification for separating banking and commerce: that moral hazard, resulting from mispriced deposit insurance, provides an incentive for the parent company to hold risky assets in the bank rather than in its nonbank subsidiaries. The argument is that because access to insured deposits provides banks with a lower cost of funds than nonbank firms, banks are more willing to hold riskier assets. However, nonbank firms will be able to sell those riskier assets to banks, whether they are affiliated or not. Thus, separating banking from commerce does not prevent risky assets from being shifted to banks. This argument is not unique to banking and commerce, as it applies equally to affiliations between banks and other financial services firms, such as those permitted under GLB. For a discussion of moral hazard as it applies to deposit insurance, see Hanc (1999), 3ff.

⁵² Walter (2003) also notes that if creditors recognize the incentive to shift losses and risks to the bank and to the FDIC, they will require lower interest rates from the bank's commercial affiliate. The resulting lower cost of capital will give these firms an incentive to engage in projects that otherwise would be unprofitable.

⁵³ See FDIC (1987a), 65-69. Carns (1995) discusses corporate separateness—the ability of firewalls and supervision to protect the insured entity—in the context of a "two-window" banking system. The arguments are applicable to the more general case of banking and commerce.

⁵⁴ See, for example, Bennett and Nuxoll (2004). The regulatory and supervisory issues raised by large and complex banks will likely be considered in the context of the New Basel Capital Accord (Basel II).

⁵⁵ The enforcement of capital standards, the monitoring of loan quality and the capability of management to run the bank, reporting requirements and disclosure standards, and the use of enforcement tools such as cease-and-desist orders and civil-money penalties, are among the supervisory tools that are used to protect the insured entity from excessive risk.

⁵⁶ Carns (1995) and Shull and White (1998), 15, discuss this issue.

⁵⁷ Walter (2003), 23, notes two instances when the breaching of firewalls led to the failure of the bank. In 1953, the failure of an Illinois bank (the First State Bank of Elmwood Park) resulted from shifts of bad loans from a nonbank loan company affiliate to the bank. In 1976, the failure of Hamilton National Bank and Trust Co., Atlanta, GA, was also caused by transactions between the bank and its affiliates.

⁵⁸ See Shull and White (1998), 9.

⁵⁹ Walter (2003), 12.

⁶⁰ Whereas both types of supervision involve a top-down approach, the former applies stricter product, service and capital requirements on the banking organization. See Kushmeider (2004).

⁶¹ GLB defined an initial set of permissible activities considered financial in nature, including securities and insurance activities and merchant banking. Additional activities for financial holding companies and their affiliates are determined by the Federal Reserve in consultation with the Secretary of the Treasury. These include activities that are financial in nature or incidental to such financial activities. In addition, financial holding companies are permitted to engage in activities that the Federal Reserve determines are complementary to financial activities, as long as the complementary activities do not pose a substantial risk to the safety and soundness of depository institutions or the financial system generally.

⁶² See, for example, Corrigan (1987, 1991).

⁶³ Cases exist in which limited-liability law has been shown to be less than perfect. In particular, the courts have occasionally disregarded limited liability—or pierced the corporate veil—when a corporation has been shown to have engaged in conduct such that creditors were led to understand that the shareholder was the true debtor. Certain safeguards can be applied to ensure that the bank and its affiliates are viewed as separate; they include separate management and record keeping for the bank and any affiliates, and boards of directors that are not identical. See, for example, Black, Miller, and Posner (1978), Walter (1996), and FDIC (1998).

⁶⁴ For a comparison of powers available to bank regulators versus holding company regulators see West (2004).

⁶⁵ Among the powers of the Federal Reserve as the umbrella supervisor of financial holding companies are the following: the power to examine the holding company; to require certain reports; to set consolidated capital standards for the banking organization, except with respect to certain functionally regulated subsidiaries; to take enforcement actions under Section 8 of the Federal Deposit Insurance (FDI) Act (12 U.S.C. §1818)); to limit activities for the holding company; to limit transactions between the insured entity and its affiliates under Sections 23A and 23B of the Federal Reserve Act (12 U.S.C. §§371(c)-371(c-1); to prohibit tying arrangements under Section 106 of the BHCA (12 U.S.C. §1971); to limit the establishment or acquisition of additional depository institutions or other subsidiaries; to disapprove changes in control; to disapprove a merger of two holding companies; to require prompt corrective action under the FDI Act; to impose cross-guarantee liability; and to prohibit golden-parachute payments.

⁶⁶ See Section 111 of GLB. The Federal Reserve has the authority to require the holding company to submit reports, but it must rely on reports made by the holding company to the functional regulator and must request unusual reports through the functional regulator. Similarly, the Federal Reserve has the authority to examine a functionally regulated subsidiary of a financial holding company, but it must first find that there is a reasonable cause to believe either that the subsidiary was engaged in activities that posed a material risk to an affiliated insured depository institution or (on the basis of reports and other available information) that the regulated subsidiary is not in compliance with relevant laws, such as Sections 23A and 23B of the Federal Reserve Act. In addition, only by examining the affiliated depository institution or its holding company can it determine that it cannot either assess that risk or decide whether a law has been violated. In establishing capital requirements, the Federal Reserve is generally precluded from imposing capital requirements on functionally regulated subsidiaries of the holding company.

⁶⁷ Edwards (1996), 161.

⁶⁸ See Carns (1995), 7-9, and Qua (2003).

⁶⁹ See U.S. Department of the Treasury (1991).

⁷⁰ Specifically, the ability to engage in expanded activities through financial affiliates would have required that the bank be well capitalized and that nonbank financial affiliates be separately capitalized. Regulation would have been focused on protecting the bank rather than on protecting the holding company or its nonbank affiliates. Nonbank affiliates would have been subject to functional regulation, and funding and disclosure firewalls would have been enforced. Responding to the criticisms that it would be harder to regulate banks if they were owned by commercial companies or that biased allocations of credit and inappropriate concentrations of economic power could result, the study noted that those risks could be addressed without a total prohibition on affiliation, and it cited as evidence the positive experience of commercial companies that own depository institutions such as thrifts, nonbank banks, and industrial loan banks. See, U.S. Department of the Treasury (1991), 54-61.

⁷¹ See ibid., 61.

⁷² See ibid., 57.

⁷³ Board of Governors of the Federal Reserve System and U.S. Department of the Treasury (2003), 10, note that more than 600 bank holding companies have chosen to become financial holding companies, while only "several" firms that were not affiliated with a commercial bank before passage of GLB have chosen to acquire a commercial bank and become a financial holding company; some of the large securities firms that have not become financial holding companies already conduct banking activities through the ownership of bank and bank-like entities that are not subject to the BHCA.

⁷⁴ MCorp Fin., Inc. v. Bd. Of Governors, 900 F.2d 852 (5th Cir. 1990), reh'g denied, 911 F.2d 730, aff'd in part, rev'd in part on other grounds, 502 U.S. 32 (1991), on remand, 958 F.2d 615 (1992). Section 112 of GLB sets conditions on how the Federal Reserve can impose its source-of-strength doctrine on the functionally regulated nonbank affiliates in the financial holding company. See, also, Qua (2003) and Wallison (2003).

⁷⁵ Individuals who own banks are not similarly required to be a "source of strength" for their banks. That is, an individual owner is not obligated to inject capital into the bank when doing so would not prevent the bank's failure.

⁷⁶ For example, on October 2, 2003, the Federal Reserve Board approved the notice of Citigroup to engage in physical commodity trading activities on a limited basis as an activity that is complementary to the financial activity of engaging regularly as principal in commodity derivative activities. See http://www.federalreserve.gov/boarddocs/press/orders/2003/20031002/default.htm .

⁷⁷ See, for example, Wallison (2000, 2003). Shull (2002) also expresses concern about whether a separation of banking and commerce can be maintained.

⁷⁸ The time-consuming regulatory process by which activities are determined to be permissible has become an issue in the debate about banking and commerce. As an example, Shull (2002) discusses the lengthy regulatory process by which the Federal Reserve considered whether real estate brokerage would be determined to be a permissible activity.

⁷⁹ Wallison (2001, 2003) is among those making this argument.

⁸⁰ Existing thrift holding companies that (a) owned a single savings and loan or other thrift institution, (b) were in existence before May 4, 1999, and (c) continued to meet the qualified-thrift-lender test were grandfathered. However, they may not engage in any new commercial activities or transfer their right to mix banking and commerce.

⁸¹ The Savings and Loan Holding Company Act of 1967 limited ownership by a diversified holding company to one thrift institution and imposed comprehensive supervisory requirements on savings and loan holding companies. The Financial Institutions Reform, Recovery and Enforcement Act of 1989 (FIRREA [Public Law 101-73, 103 Stat. 183]) imposed tougher firewalls on unitary thrift holding companies; prohibited lending between the thrift and any affiliate that engaged in activities not permissible for a bank holding company; and subjected transactions with other affiliates to the same restrictions imposed under Sections 23A and 23B of the Federal Reserve Act. See Doyle, DeSimone and Biddle (1999).

⁸² See Shull and White (1998).

⁸³ OTS (1998), 6.

⁸⁴ See Thomson (2001).

⁸⁵ See Doyle, DeSimone and Biddle (1999).

⁸⁶ Although the Federal Reserve has consolidated supervisory authority over U.S. bank holding companies, this authority does not extend to the parent companies of foreign banks. The home-country regulator is responsible for regulating and examining the consolidated operations of the foreign bank. As a result, it can be argued that the Federal Reserve effectively uses a bank-up approach in its supervision of the operations of foreign banks in the United States. See, Foreign Banks: Assessing Their Role in the U.S. Banking System, GAO/GGD-96-26, 48.

⁸⁷ See FDIC (1987a), 91.

⁸⁸ Moreover, FIRREA provided the FDIC with the ability to recover from solvent affiliated banks the losses the FDIC has incurred from the failure of an insured bank. Two federal circuit courts of appeal have upheld the constitutionality of that cross-guarantee provision (see Walter [1996], 34). In terms of the banking and commerce debate, the cross-guarantee provision should reduce the incentive for bank owners to shift losses to the bank.

⁸⁹ 12 U.S.C. §1820(b)(4).

⁹⁰ The ILC charter has been in existence since the early 1900s. Referred to as "industrials" or "thrift and loans," ILCs typically provided banking services to industrial workers, and until the 1940s ILCs operated in most states. Today, ILCs are found mainly in Utah and California and also operate in Colorado, Hawaii, Indiana, Minnesota, and Nevada. As of December 2003, there were 56 ILCs; 73 percent of ILCs reported total assets of less the $500 million; approximately 10 percent of ILCs reported total assets greater than $5 billion. See West (2004) for further discussion of ILCs.

⁹¹ ILCs generally maintain this exemption by meeting at least one of the following conditions: (1) control of the ILC has not been acquired by any company since August 10, 1987; or (2) the ILC does not accept demand deposits; or (3) the ILC maintains total assets of less than $100 million. 12 U.S.C. §(c)(2)(H).

⁹² Leary (2003).

⁹³ Many Utah ILC approval orders impose the following key conditions: an independent board with a majority of outside, unaffiliated directors; no change in executive officers for the first three years of the ILC's operation; on-site officers, including (at a minimum) a chief financial officer and a chief credit officer with sufficient staff and the knowledge, ability, and expertise to successfully manage the ILC, maintain direct control of it, and retain its independence from the parent company; and monthly meetings of the board of directors for the first two years of the ILC's operation (Leary [2003]).

⁹⁴ Powell (2003a) and West (2004).

⁹⁵ Conseco Corporate Website (2003).

⁹⁶ Hoovers Online (2003).

⁹⁷ Green Tree Retail Services Bank, the South Dakota credit-card bank, voluntarily liquidated in 2003.

⁹⁸ Wisniewski (2003).

⁹⁹ GE Capital did not purchase the bank in toto but purchased the majority of the bank's assets and assumed all of its liabilities. As of July 26, 2003, the bank forfeited its charter and voluntarily liquidated without cost to the FDIC.

¹⁰⁰ See Kushmeider (2004), and Ferguson (2000), 2.

¹⁰¹ Walter (2003), 24.

¹⁰² Again, see U.S. Treasury (1991), FDIC (1987a) and Leary (2003).

¹⁰³ For example, problems in Japan and Korea have led to such losses. See Seidman (2003).

¹⁰⁴ Walter (1996), 34-35, argues that to reduce the likely occurrence of loss shifts, the size of commercial affiliates should be restricted. However, with sufficient firewalls and the regulator's ability to monitor both sides of any transaction and to examine the bank and its affiliates, including the parent, it can be argued that size alone should not be a reason to prohibit affiliations.

¹⁰⁵ During testimony on financial services reform, the FDIC (1997) noted that in light of the increasing complexity of the financial marketplace, some form of added oversight of banks and other providers of financial services could enhance coordination and attention to interstitial concerns, such as maintaining accurate information about all operations in the organization and monitoring compliance with the rules on inter-company dealings. The FDIC further noted that for safety-and-soundness purposes, it would not be necessary "to include investment-by-investment or activity-by-activity regulation as part of the oversight of the consolidated organization, provided that risks to the financial system and to the insurance funds are understood and appropriately limited" (p. 10).

¹⁰⁶ Kushmeider (2004) discusses the use of complexity as the criterion for determining appropriate regulatory oversight.

¹⁰⁷ Policy makers should also consider the need for improved communications between the banking organization and the regulatory authorities, for which both parties must take responsibility. Bankers are responsible for understanding what takes place in their organizations and for communicating this information to their regulators. Their incentive, of course, is that the possible alternative is additional regulation. Similarly, the regulatory authorities must clearly communicate relevant information about the regulations and supervisory procedures; the goal would be better mutual understanding through discussions and communications with the boards of directors and corporate managers. Communication between functional regulators in a complex institution can also contribute to protecting the institution and, more broadly, the safety net. An example of such communication is the cooperation between the FDIC and the Utah Department of Financial Institutions in supervising ILCs.

¹⁰⁸ See Powell (2004).

REFERENCES

Barth, James R., Gerard Caprio, Jr., and Ross Levine. 2000. Banking Systems around the Globe: Do Regulation and Ownership Affect Performance and Stability? Working Paper No. 2325. World Bank.

Barth, James R., Gerard Caprio, Jr., and Daniel E. Nolle. 2004. Comparative International Characteristics of Banking. Economic and Policy Analysis Working Paper 2004-1. Office of the Comptroller of the Currency.

Bennett, Rosalind L. and Daniel A. Nuxoll. 2004. Large Bank Regulatory Issues: Supervision, Market Discipline and Capital Standards. Unpublished manuscript. FDIC.

Black, Fischer, Merton H. Miller, and Richard A. Posner. An Approach to the Regulation of Bank Holding Companies. Journal of Business 51:379-412.

Blair, Christine E. 1994. Bank Powers and the Separation of Banking and Commerce: An Historical Perspective. FDIC Banking Review 7, no. 1:28-38.

____________. 2004. The Separation of Banking and Commerce: Evidence Versus Myth. Unpublished manuscript. FDIC.

Board of Governors of the Federal Reserve System and U.S. Department of the Treasury. 2003. Report to the Congress on Financial Holding Companies under the Gramm-Leach-Bliley Act.

Carns, Frederick S. 1995. A "Two-Window" System for Banking Reform. FDIC Banking Review 8, no. 2:1-12.

Conseco Corporate Website. 2003. Conseco History. http://www.conseco.com/conseco/selfservice/about/cprofile/history.jhtml?cat=cprofile&subcat=hist [May 2004].

Corrigan, E. Gerald. 1987. Financial Market Structure: A Longer View. Federal Reserve Bank of New York.

____________. 1991. The Banking-Commerce Controversy Revisited. Federal Reserve Bank of New York Quarterly Review 16, no. 1:1-13.

Craig, Valentine V. 2004. Bank Governance: Issues Arising from Investment Banking Activities. Unpublished manuscript. FDIC.

Doyle, A. Patrick, Beth S. DeSimone and Laura Biddle. 1999. Unitary Thrift Holding Company Worry is Overdrawn. The National Law Journal 21, no. 37. http://www.arnoldporter.com/publications_articles.cfm?practice_ID=0&publication_id=329 [August 2004].

Edwards, Franklin R. 1996. The New Finance: Regulation and Financial Stability. AEI Press.

Federal Deposit Insurance Corporation (FDIC). 1987a. Mandate for Change: Restructuring the Banking Industry. FDIC.

____________. 1987b. Testimony of L. William Seidman, Chairman, FDIC, on Financial Services Reform Legislation before the U.S. Senate Committee on Banking, Housing and Urban Affairs. 100th Cong., 1st sess., December 3.

____________. 1997. Testimony of Ricki Helfer, Chairman, FDIC, on Financial Modernization, before the U.S. House Committee on Banking and Financial Services, Subcommittee on Capital Markets, Securities and Government Sponsored Enterprises. 105th Cong., 1st sess., March 5. http://www.fdic.gov/news/news/speeches/archives/1997/sp05mar97.html [October 2003].

____________. 1998. Testimony of Donna Tanoue, Chairman, FDIC, on Financial Modernization, before the U.S. Senate Committee on Banking, Housing, and Urban Affairs. 105th Cong., 2nd sess., June 25. http://www.fdic.gov/news/news/speeches/archives/1998/sp25june98.html [May 2004].

Ferguson, Roger W., Jr. 2000. Umbrella Supervision: Emerging Approaches. Remarks before the National Association of Urban Bankers, Urban Financial Services Coalition. San Francisco, CA. May 26. http://www.federalreserve.gov/boarddocs/speeches/2000/20000526.htm [April 12, 2004].

Golembe, Carter H. 1997. Separation of Banking and Commerce: A Myth That's Ripe for Debate. Banking Policy Report 16, no. 1:12-17.

Greenspan, Alan. 2003. Letter to Senator Tim Johnson, June 25. http://lib1/division/doa/library/hotdocs/07162003/greenspan-ilc.pdf [April 1].

Halpert, Stephen K. 1988. The Separation of Banking and Commerce Reconsidered. Journal of Corporation Law 13 (Winter): 481-533.

Hammond, Bray. 1936. Free Banks and Corporations: The New York Free Banking Act of 1838. Journal of Political Economy 44 (February): 184-209.

____________. 1957. Banks and Politics in America from the Revolution to the Civil War. Princeton University Press.

Hanc, George. 1999. Deposit Insurance Reform: State of the Debate. FDIC Banking Review 12, no. 3:1-26.

Haubrich, Joseph G., and João A. C. Santos. 1999. Banking and Commerce: A Liquidity Approach. Working Paper No. 78. Bank for International Settlements.

____________. 2003. Alternative Forms of Mixing Banking and Commerce: Evidence from American History. Financial Markets, Institutions & Instruments 12, no. 2:121-59.

Hoovers Online. 2003. Conseco Finance Corp. Hoovers Company Profiles.

Jones, Kenneth, and Barry Kolatch. 1999. The Federal Safety Net, Banking Subsidiaries, and Implications for Financial Modernization. FDIC Banking Review 12, no. 1:1-17.

Jorde, Terry J. 2003. Comments at the FDIC Symposium The Future of Banking: The Structure and Role of Commercial Affiliations. Washington, D.C., July 16. http://www.fdic.gov/news/conferences/future_jordeSpeech.html [July 22].

Klebaner, Benjamin J. 1990. American Commercial Banking: A History. Twayne Publishers.

Krainer, John. 2000. The Separation of Banking and Commerce. Federal Reserve Bank of San Francisco Economic Review, 15-25.

Kushmeider, Rose M. 2004. The U.S. Federal Financial Regulatory System: Restructuring Federal Bank Regulation. Draft FOB-2004-12.1. FDIC.

Leary, G. Edward. 2003. Opening Statement. Remarks before the FDIC Symposium on The Future of Banking: The Structure and Role of Commercial Affiliations. Washington, D.C., July 16. http://www.dfi.state.ut.us/PDFiles/FDIC%20Symposium-Ed%20Leary.pdf .

Office of the Comptroller of the Currency (OCC). 2003. Today's Credit Markets, Relationship Banking, and Tying. OCC.

Office of Thrift Supervision (OTS). 1998. Holding Companies in the Thrift Industry. OTS. Powell, Donald. 2003a. Letter to Senator Robert Bennett. April 30.

____________. 2003b. The ILC Debate: Regulatory and Supervisory Issues. Remarks before the Conference of State Bank Supervisors, May 30. http://www.fdic.gov/news/news/press/2003/pr5203.html [June 2].

____________. 2004. Remarks before the National Association for Business Economics, March 26. http://www.fdic.gov/news/news/press/2004/pr3004.html [April 8].

Pozdena, Randall J. 1991. Why Banks Need Commerce Powers. Federal Reserve Bank of San Francisco Economic Review 3:18-31.

Qua, John. 2003. Comments at the FDIC's Symposium The Future of Banking: The Structure and Role of Commercial Affiliations. Washington, D.C., July 16. http://www.fdic.gov/news/conferences/future_qua_speech.html [July 22].

Santos, João A. C. 1998. Banking and Commerce: How Does the United States Compare to Other Countries? Federal Reserve Bank of Cleveland Economic Review 34:14-26.

____________. 1999. Bank Capital and Equity Investment Regulations. Journal of Banking and Finance 23:1095-1120.

Saunders, Anthony. 1994. Banking and Commerce: An Overview of the Public Policy Issues. Journal of Banking and Finance 18:231-54.

Seidman, L. William. 2003. Comments at the FDIC Symposium, The Future of Banking: The Structure and Role of Commercial Affiliations. Washington, D.C., July 16. http://www.fdic.gov/news/conferences/future_transcript.html#Seidman [July 22].

Shull, Bernard. 1999. The Separation of Banking and Commerce: An Examination of the Principal Issues. E&PA Working Paper 99-1. Office of the Comptroller of the Currency.

____________. 2002. Banking, Commerce and Competition under the Gramm-Leach-Bliley Act. Antitrust Bulletin 47, no. 1:25-61.

Shull, Bernard, and Lawrence J. White. 1998. Of Firewalls and Subsidiaries: The Right Stuff for Expanded Bank Activities. Working Paper CLB-98-017. Center for Law and Business, New York University.

Symons, Edward L., Jr. 1983. The "Business of Banking" in Historical Perspective. The George Washington Law Review 51, no. 5:676-726.

Thomson, James B. 2001. Unitary Thrifts: A Performance Analysis. Federal Reserve Bank of Cleveland Economic Review 37, no. 2:2-14.

U.S. Department of the Treasury. 1991. Modernizing the Financial System: Recommendations for Safer, More Competitive Banks. U.S. Department of the Treasury.

U.S. General Accounting Office. 1996. Foreign Banks: Assessing Their Role in the U.S. Banking System. GOA-GGD-96-26.

____________. 2003. Bank Tying: Additional Steps Needed to Ensure Effective Enforcement of Tying Prohibitions. GAO-04-4.

Wallison, Peter J. 2000. The Gramm-Leach-Bliley Act Eliminated the Separation of Banking and Commerce: How This Will Affect the Future of the Safety Net. Presentation at the Federal Reserve Bank of Chicago's Annual Conference on Bank Structure and Competition. May 5.

____________. 2001. Why Has the Gramm-Leach-Bliley Act Failed? Introduction to AEI Event on Banking Reform. November 13. http://www.aei.org/news/filter.all,newsID.15984/news_detail.asp [July 8, 2003].

____________. 2003. The Gramm-Leach-Bliley Act Eliminated the Rationale for the Separation of Banking and Commerce. Comments at the FDIC's Symposium The Future of Banking: The Structure and Role of Commercial Affiliations. Washington, D.C. http://www.aei.org/news/newsID.17994,filter./news_detail.asp [July 16].

Walter, John R. 1996. Firewalls. Federal Reserve Bank of Richmond Economic Quarterly 82, no. 4:15-39.

____________. 2003. Banking and Commerce: Tear down This Wall? Federal Reserve Bank of Richmond Economic Quarterly 89, no. 2:7-31.

West, Mindy. 2004. The FDIC's Supervision of Industrial Loan Companies: A Historical Perspective. FDIC Supervisory Insights 1, no. 1:5-13.

Wisniewski, Mary. 2003. Conseco Finance Sale to CFN, GE Approved by Judge. Bloomberg News. [March 14, 2004].