sidiaries and not in bank holding company affiliates, when they are permitted to choose between those two options. Again, bank behavior is not consistent with the presence of a subsidy. For example, banks can locate their mortgage banking operations in a bank, a bank subsidiary, or in an affiliate of a holding company. Of the top twenty bank holding companies, six conduct mortgage banking operations in a holding company affiliate, nine conduct mortgage banking activities in the bank or bank subsidiaries, and five conduct mortgage lending through a combination of the bank and holding company. The table below lists other activities such as consumer finance, leasing, and data processing-that banking companies offer through both holding company affiliates and bank subsidiaries.

Most Common Nonbank Affiliates of Bank Holding Companies and
Subsidiaries of Banks: 1996 6

In offering many of the activities shown in the table above, banks compete sideby-side with nonbank providers. If banks had a competitive advantage, they would dominate over other providers. However, in many fields, nonbank providers have a bigger market share than banks. As of December 1996, three out of the top five largest servicers of residential mortgages were nonbanks, and three of the top five originators of mortgages were nonbanks.8 The Federal Reserve has stated persuasively that banks engaging in permissible securities activities do not dominate their respective markets.9

Evidence Offered To Support The Subsidy Claim Does Not Withstand Scrutiny

The five pieces of evidence cited in support of the existence of a subsidy are: (1) bank debt is rated higher than that of its parent bank holding company, (2) banks hold less capital than other financial institutions, (3) bank capital ratios declined in the twentieth century, (4) corporations are not leaving the banking business; and (5) bank holding companies are shifting activities from affiliates to banks or bank subsidiaries. In fact, none of these points demonstrates the presence of a safety net subsidy.

First, the small differential between the ratings of debt issued by banks and debt issued by bank holding companies is not due to a safety net subsidy. In 1996, this rating differential resulted in a cost of funds for bank holding companies that was only 4 to 7 basis points higher than the cost of funds for individual banks. According to the rating agencies, the difference is due to the Federal banking agencies' ability to use prompt corrective action powers to limit bank payments to the holding company if the bank is undercapitalized. A bank holding company is a shell corporation,

6 Data as of September 30, 1996. Includes all direct subsidiaries of the bank or holding company. All banks in this analysis were members of holding companies. Source: Federal Reserve Board National Information Čenter. 7 Insurance agency or brokerage services related to credit insurance. 8"Ranking the Banks: Statistical Review 1996," American Banker.

In its 1987 ruling, "Order Approving Activities of Citicorp, J.P. Morgan, and Bankers Trust to Engage in Limited Underwriting and Dealing in Certain Securities, Legal Developments," the Federal Reserve Board stated, "the Board notes that banks do not dominate the markets for bank-eligible securities, suggesting that the alleged funding advantages for banks are not a significant competitive factor."

with most of its assets held by, and income generated by, the subsidiary bank(s). Reductions in the flow of funds from the banks to the corporate shell decreases the debt-paying capacity of the holding company parent.

A second fact cited to support the existence of a safety net subsidy is that banks hold less capital than virtually all other financial institutions. This argument is flawed because it makes no sense to compare capital ratios of different industries in isolation from their relative risk. For example, two institutions engaged in very different lines of business could have distinct risk profiles. The market would demand a higher equity-to-assets ratio of the firm that holds much riskier assets in its portfolio. Merely comparing the institutions' capital ratios is insufficient, and a finding that banks' ratios are lower does not prove that there is a subsidy.

Moreover, accounting approaches vary across industries. For example, banks tend to use historical cost reporting, whereas securities firms use mark-to-market accounting. Such accounting differences might explain differences in reported capital ratios between the two industries. Consequently, any attempt to attribute the dif ferences in reported capital ratios to the alleged subsidy must account for these differences in accounting practices.

Third, the decline in bank capital ratios in the decades following the creation of the Federal Reserve System and the FDIC did not result from a safety net subsidy. A more plausible explanation is that capital ratios declined because the efficiency of the U.S. financial system has increased over time. A 1991 Treasury Department study concluded that "[c]apital ratios were declining long before creation of either the Federal Reserve System or the FDIC. Indeed, much of the decline both before and after the creation of the safety net no doubt reflects the growing efficiency of the U.S. financial system.” 10 A copy of a chart from that study showing the decline in equity as a percent of assets for all insured commercial banks from 1840 to 1989 is attached as Appendix 1.

Fourth, some banking industry observers have argued that the fact that corporations are not leaving the banking business is evidence that a subsidy exists. However, subsidy proponents must also explain why the alleged subsidy in banking has not attracted other firms. One possible explanation is that there are barriers to entry; yet banking is a highly-regulated industry. This counter argument cannot end here. If there were substantive barriers to entry and no other factors were at work, banks should experience excessive profits and a growing niarket share. The facts are not consistent with those implications. Bank profits, while strong in recent years, are not disproportionately higher than other competitors in the financial services industry.11 Bank stock price-to-earnings (P/E) ratios have, on average, averaged only about 60 percent of P/E ratios of other businesses. 12 Also, banks' market share, measured by income-based data, has remained flat at least since the late 1950's.13 Moreover, subsidy proponents must also explain why industry consolidation, which is a form of exiting from banking, is not at odds with their view of the facts.

Finally, those who are seeking to prove the existence of a subsidy cite more recent developments as evidence. In particular, they point to a reported drop over the last decade in the share of bank holding company assets held by nonbank subsidiaries, after removing the Section 20 affiliates (firms engaged in Federal Reserve approved securities activities). The argument seems to be that such a shift is motivated by a desire to exploit a subsidy available to banks and their subsidiaries but unavailable to affiliates of bank holding companies. It is one thing, however, to make that observation, and another to conclude that it is due to a subsidy.

First, there are alternative explanations for banking organizations moving activities from holding company affiliates to banks. Importantly, over the past decade, the relaxation of geographical and other barriers to interstate banking has permitted banking companies to engage in the interstate conduct of lines of business in banks which they could previously conduct only through holding company subsidiaries. That flexibility could lead banking organizations to shift assets from long-established holding company subsidiaries in those States to newly available banks or

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

11 According to data presented in the Property/Casualty Fact Book 1997 published by the Insurance Information Institute, banks had a lower annual rate of return than diversified financial services firms for all but 2 years in the period 1986 through 1995, the last year for which comparable data are available. However, as is true when comparing capital ratios, it is difficult to make a direct comparison of profits without making a risk-adjustment. In other words, it is difficult to determine whether profits are commensurate with risks undertaken.

12 According to data by Keefe, Bruyette & Woods, Inc., commercial bank P/E earnings as a percentage of the S&P 500 P/E ratio averaged 62 percent for the 6 years ending April 15, 1997. 13 George Kaufman and Larry Mote, "Is Banking a Declining Industry? A Historical Perspective," Economic Perspectives, Federal Reserve Bank of Chicago (May/June 1994), pp. 2-21.

bank subsidiaries. Moreover, firms consolidate their operations for many reasons, including the desire for increased efficiency. Recent experience with intrastate and interstate branching demonstrates the efficiency gains of organizational flexibility. Research on intracompany mergers finds that choice of organizational form is an important determinant of the efficiency of a company's operations. These mergers enable banking organizations to streamline their operations and better serve their customers.14 After many States eased restrictions on intrastate branching, most banking companies responded by consolidating all of their existing subsidiaries into branch banks, although this was not the universal response.

15

Second, we cannot be sure that an asset shift occurred. There are no systematic data available to document that a shift occurred, and unless we have such data, we must reserve judgment on any implications they may hold for the subsidy debate. The existing data are problematic for several reasons: between 1994 and 1995, the Federal Reserve changed the instructions governing the filing of the asset data used in the calculation of the reported shift to reduce, if not eliminate, apparently the widespread, year-by-year, reporting errors. The presence of these reporting errors and the changes in reporting instructions mean that we cannot make accurate yearto-year comparisons. Indeed, the absence of comparability could fully account for the reported drop in holding company affiliate share of bank holding company assets. In sum, there is a better, alternative explanation for every piece of evidence cited by observers who believe there is a subsidy. Mandating organizational form because of an alleged subsidy is not good public policy because it ignores important facts. The existence of a gross benefit is insufficient to prove that banks have an unfair competitive advantage. What matters is whether banks have a benefit after netting out the costs of that benefit-a net subsidy and the evidence is that they do not. Any benefit to banks from access to the safety net has declined significantly over the past decade. Conservative current estimates of regulatory costs imply that the net subsidy received by banks is negative. Most important, banks do not behave as if they were subsidized.

III. Organizational Structure Cannot By Itself Determine The

Beneficiaries Of Any Subsidy; Rules Can. Current Rules Make The
Bank Operating Subsidiary Structure Superior To The Holding
Company Structure In Keeping Any Subsidy Derived From The
Federal Safety Net Where It Belongs-In The Bank

Sound public policy requires that we strive to keep the benefits of any subsidy, if it exists, within the bank. However, mandating that banking companies create one corporate form will not achieve that objective because organizational structure by itself is not enough to confine the benefits of a subsidy. Rules governing the transfer of funds among affiliated entities are necessary. Such rules could ensure that operating subsidiaries and bank holding company subsidiaries are equally effective at confining the subsidy to the bank.

Under existing rules, transmission of any subsidy in the form of dividends from the bank through the parent holding company to affiliates would be relatively easy. The transmission from the bank to a bank operating subsidiary 16 would be more difficult.

The ability to confine any subsidy would depend not on where we place new activities in the financial organization chart, but on what restrictions we impose on funds transfers between a bank and its subsidiaries and other affiliates. Under current rules, the bank subsidiary structure may better prevent any subsidy from flowing beyond the bank. The OCC's part 5 regulation imposes the same limitations on transactions between a bank and a bank subsidiary engaged, as principal, in an activity not permitted for the bank as those applied by Sections 23A and 23B of the Federal Reserve Act to transactions between a bank and its holding company affiliates. (A bank may invest no more than 10 percent of its capital in a subsidiary.) Furthermore, the OCC's regulation permits only well-capitalized banks to make such investments, and the bank must remain well capitalized for regulatory purposes after deducting the equity investment. In other words, the regulation explicitly restricts the amount of funds a bank may downstream to a subsidiary.

14 Robert DeYoung and Gary Whalen, "Is a Consolidated Banking Industry a More Efficient Banking Industry," OCC Quarterly Journal, September 1994.

15 Robert DeYoung and Gary Whalen, "Banking Industry Consolidation: Efficiency Issues," Working Paper No. 100, The Jerome Levy Economics Institute, April 1994.

16 In this section, any reference to a bank operating subsidiary refers to a subsidiary engaged in an activity not permitted for the bank as principal.

Any subsidy can be passed up to the holding company through the payment of dividends, and an adequately capitalized bank faces few formal restrictions when paying dividends to its bank holding company parent. Neither Sections 23A and 23B of the Federal Reserve Act nor any comparable restrictions apply to payment of dividends by a bank to its holding company, and banks need only to be adequately capitalized to pay dividends. As long as earnings are adequate, there is no limit on the amount of funds an adequately capitalized bank can upstream to a holding company affiliate. Because most banks today are well capitalized, banks could help holding company affiliates provide approved activities simply by paying dividends.

Once a dividend is paid, it is impossible to determine whether that cash flow goes to affiliates or bank holding company shareholders. Although proponents of the holding company affiliate approach acknowledge that bank dividends flow upward, they claim that bank profits do not fund bank holding company affiliates. The truth is elusive. Since money is fungible, no one can pinpoint the role dividends play in affiliate operations. Moreover, even if bank dividends played little if any role in the past, they could still play a substantive role in the future. As financial modernization increases the range of activities that bank holding companies can conduct, there will be greater incentive for those companies to use bank profits to fund activities of their nonbank affiliates. Limits on dividend payments would be necessary to confine any alleged subsidy to the bank.

Rules are necessary to prevent the transfer of any alleged subsidy. Under current rules, the bank subsidiary structure is stronger than the bank holding company structure in confining any subsidy. The bank subsidiary struc. ture is also a sound framework for supporting the continued safe and sound operation of our Nations' banking system.

IV. The Logical Consequence Of Organizational Rigidity-The Destabilized Hollow Bank-Is Unacceptable

The only sure way to prevent transmission of the alleged subsidy would be to reject financial modernization altogether and to limit banks and all of their affiliated companies to a narrow range of activities. Permitting holding company nonbanks to offer services bank subsidiaries could not supply would not confine the transmission of any subsidy. Instead, such limitations on bank subsidiaries would deprive banks of important sources of income, yielding a destabilized hollow bank. Both alternatives are unacceptable.

Rejecting financial modernization altogether and limiting bank subsidiaries are unacceptable options because both would deprive banks, their customers, and the general economy of the important benefits of modernization. By contrast, allowing banks to conduct new activities in operating subsidiaries could have positive benefits while limiting downside risks.

The only way to guarantee that any alleged subsidy will not be transmitted would be to limit the activities of banking companies. Proponents of the holding company structure assert that there would be no way to prevent at least some benefit associated with any alleged subsidy from leaking from banks to holding company parents and affiliates. If they are correct and if this benefit should not be available outside the bank, the logical conclusion would be to reject financial modernization altogether and to limit banks and all of their affiliated companies to a narrow range of activities.

Permitting holding company nonbanks to offer activities denied bank subsidiaries would not confine the transmission of any subsidy. Instead, such limits would lead to a destabilized hollow bank. Forcing a bank to offer new products and services only through a holding company affiliate will limit the bank's ability to respond to changes in the marketplace and impose unnecessary costs that will render the bank less competitive. Either the assets and income stream of the bank itself will shrink, or the bank will feel pressure to reach ever farther out on the risk curve to be profitable and generate adequate returns on capital and to remain in business. The result will be a destabilized hollow bank that is less safe and sound, offers fewer choices to customers, and is less able to serve our communities and the broader financial needs of its customers.

Moreover, organizational rigidity may deprive some communities of competitive alternatives. Imposing a cumbersome holding company structure on small banks wishing to conduct critical activities may create costs and inefficiencies that make these activities unprofitable for them. Such costs may even mean that residents of some communities may not have a local supplier of some valued financial services and that residents of other communities may find that the number of competitors is less than it could be.

By contrast, allowing banks to conduct new activities in operating subsidiaries could have positive benefits while limiting downside risks. There are benefits from

banks that diversify earnings by conducting some activities through operating subsidiaries. Fees and other income from the subsidiaries may enable banks to offset the effects of cyclical downturns in other sectors of the economy. Hence, bank earnings would be less volatile, reducing risks to the banking system as a whole.17 Draws on the deposit insurance fund from bank failures will be less likely, in part, because the assets of bank operating subsidiaries are clearly available to the FDIC.18 Stronger banks will be able to make more credit available to the economy. Structural flexibility gives banks the ability to serve the evolving needs of their customers, improve efficiencies, and therefore effectively compete at home and abroad. Recent experience with intrastate and interstate branching demonstrates the efficiency gains and consumer benefits of organizational flexibility. OCC research on intracompany mergers in banking finds that choice of organizational form is an important determinant of the efficiency of a company's operations. 19

Organizational flexibility is also critical to ensuring fair access to financial services. Forcing activities out of the bank reduces the resources available to support the bank's Community Reinvestment Act (CRA) efforts. By contrast, allowing a bank subsidiary to engage in a wide range of activities keeps earnings flowing up to the bank and available for funding CRA initiatives. The regulators consider the assets of a bank subsidiary when they assess the capacity of the bank to serve its community.20

Bank subsidiaries can only be a source of strength to banks, not a source of weakness. Subsidiary earnings and assets are always available to the bank. The reverse is not true. If a bank subsidiary becomes troubled, bank regulators can force the bank to divest itself of the subsidiary under the prompt corrective action provisions of FDICIA. When this occurs, the parent bank's financial statements reflect a loss equal to the loss in value of the bank's total investment in the subsidiary. The bank's total investment would include any guarantees or commitments to the subsidiary, to the extent the bank's regulator allows such guarantees or commitments. They would not reflect the losses incurred by the subsidiary because the subsidiary is a separate corporation. The result is a financial structure that enhances the safety and soundness of the banking system.

The evidence cited above demonstrates that organizational rigidity is not only unnecessary but also counterproductive. There is no sound public pol icy reason to limit a bank's range of financial activities or to impose a par ticular corporate structure for conducting those activities. There is no net subsidy from the safety net, nor does the holding company structure better confine any subsidy. Rather, in the interests of safety and soundness for our Nation's banking system, it is incumbent upon policymakers to: (1) change the rules governing the transmission of funds among banking affiliates to make bank and holding company subsidiaries equally effective in containing any subsidy, and (2) choose a financial modernization path that embraces organizational flexibility.

V. Summary and Conclusions

There is no reason to restrict bank organizational structure. There is no net subsidy from access to the safety net. Gross benefits to banks have decreased in recent years, and regulatory costs offset those benefits. The holding company structure is not proof against the leakage of any alleged subsidy because banks could pass that subsidy upstream to the holding company through dividends, and an adequately capitalized bank's ability to pay dividends is subject to few restrictions. At the same time, a bank's investment in a subsidiary is explicitly limited, and earnings diversification from such an investment may improve safety and soundness.

Any concern that banks have a funding advantage that should not be used in the marketplace can only be resolved completely by rejecting financial modernization al

17 See for example, Peter S. Rose, "Diversification of the Banking Firm," The Financial Review, vol. 24 (May 1989), pp. 251-280.

18 See testimony of Ricki Helfer, Chairman, FDIC, on financial modernization before the Subcommittee on Capital Markets, Securities, and Government-Sponsored Enterprises, Committee on Banking and Financial Services, U.S. House of Representatives, March 5, 1997.

19 These mergers enable banking organizations to streamline their operations and better serve their customers. Robert DeYoung and Gary Whalen, "Is a Consolidated Banking Industry a More Efficient Banking Industry," OCC Quarterly Journal, September, 1994.

20 As Allen Fishbein, General Counsel of the Center for Community Change, recently noted, it is also important to understand that the OCC's new rule provides a potentially important means for increasing the resource base for CRA-related activities." See written testimony, hearing before the Subcommittee on Financial Institutions and Consumer Credit, Committee on Banking and Financial Services, U.S. House of Representatives, February 25, 1997.