been no financing of a bank's holding company affiliates with subsidized equity of the associated bank.

The dividend flows from banks to their parent holding companies have been less than the sum of holding company dividends, interest on holding company debt, and the cost of holding company stock buybacks, a substitute for dividends. All of that part of the subsidy reflected in earnings has flowed directly to investors.

That bank dividends are not used to finance holding company subsidiaries should not be surprising. It simply is not in the interest of the consolidated banking organization to increase bank dividend flows beyond parent company capital-servicing needs because the resulting decline in bank capital would increase funding costs of the bank.

Research at the Federal Reserve indicates that over the past quarter century, for the largest banks the cost of uninsured bank funds has tended to rise as a bank's capital ratio fell and viceversa. This is just what one should expect. As the risk-absorbing equity cushion falls, the risk for uninsured creditors rises. The flow of dividends from the bank to the parent holding company reduces bank capital. That reduction in turn reduces the risk buffer for uninsured creditors, increasing the funding costs of the bank on all the uninsured liabilities by more, the data show, than the small subsidy transference of funding the additional equity investment in the affiliate.

Thus, were a bank holding company to finance its nonbank affiliates from bank dividends, that is, to directly pass on the bank subsidy to the holding company affiliates, the profitability of the consolidated organization would decline. If there were no net costs to the bank from upstreaming dividends to its parent for affiliate funding, it would be the prevalent practice today. It is not. In short, the subsidy appears to have been effectively bottled up in the bank. The Federal Reserve Board believes that this genie would be irreversibly let out of the bottle, however, should the Congress authorize wider financial activities in operating subs. Subsidized equity investments by banks can be made in their own subsidiaries without increasing funding costs on all of the bank's uninsured liabilities because the consolidated capital of the bank would not change in the process. When a bank pays dividends to its parent, the bank shrinks, and its capital declines. When a bank invests in its subsidiary, its capital remains the same.

None of this is relevant today since the activities authorized to bank subsidiaries cannot differ from those available to the bank itself under current law. Hence, there is no additional profit to the overall banking organization in shifting existing bank powers to a subsidiary-the activity would receive the same subsidy in the subsidiary as it now gets in the bank. But H.R. 10 would currently permit activities not now permitted in the bank. Those activities, when performed in bank subsidiaries and financed with subsidized bank equity capital, would increase the potential profit to the overall banking organization. It would also inevitably induce the gravitation to subsidiaries of banks, not only of the new powers authorized by H.R. 10, but all of those powers currently financed in holding company affiliates at higher costs of capital than those available to the banks.

How important is this subsidy? Even today when losses in the financial system and hence the value of the subsidy are quite low, the cost of debt capital to banks still averages 10 to 12 basis points below that of the parent holding companies. That difference in bond ratings today between banks and the holding companies, let alone the larger difference between banks and other financial institutions, is a significant part of the 20 to 30 basis point gross margin on an A-rated or better investment grade business loans, more than enough to significantly change lending behavior if it were not available.

Business loan markets are particularly competitive, and hence there is little leeway for a competitor with higher funding costs to pass on such costs to the borrower. For example, the weakened credit standing of the Japanese banks has engendered a risk premium that these entities have paid and today would have to pay to fund their U.S. affiliates. This has required them to sharply reduce their business loan volume in the United States. Japanese bank branches and agencies in the United States have reduced their share of business loans from over 16 percent of the total U.S. market in 1995 to less than 11 percent today.

In short, the subsidy is a critical competitive issue in competitive markets. Allowing the bank to inject Federal subsidies into the proposed new activities could distort capital markets and the efficient allocation of both financial and real resources. New affiliations, if allowed through banks, would accord them an unfair competitive advantage over comparable nonbank firms. The holding company structure, on the other hand, fosters a level playing field within the financial services industry contributing to a more competitive environment.

Mr. Chairman, in addition to our concern about the extension of the safety net that would accompany the_widening of bank activities through operating subsidiaries, the Federal Reserve Board is also sensitive to the implications of operating subsidiaries for the safety and soundness of the parent bank. Most of the new activities contemplated by H.R. 10 would not be accompanied by unusually high risk, but they could imply more risk. Although, to be sure, diversification can reduce that risk, the losses that would accompany riskier activities from time to time would fall on the insured bank's capital if the new activities were authorized in bank subsidiaries. Such losses at holding company affiliates would fall on the uninsured holding company. This is an important distinction for the deposit insurance funds and potentially the American taxpayer. This potential for loss and bank capital depletion is another reason for urging that the new activities be conducted in a holding company affiliate rather than in a banking subsidiary.

H.R. 10 is supposed to virtually eliminate this concern. The Office of the Controller of the Currency has asserted that it would order an operating sub immediately to be sold or declared bankrupt and closed before its cumulative losses exceeded the bank's equity investment in the failing sub. Combined with the provision of H.R. 10 adjusting regulatory capital for investment in subs, this provision is intended to cap the effect on the bank of subsidiary loss to the amount of the bank's original investment. Since that amount would have already been deducted from the bank's regulatory cap

ital, the failure of a subsidiary, it is maintained, could not affect the regulatory capital of the bank.

We had extensive experience with attempts to redefine reality by redefining regulatory capital in the thrift industry in the 1980's. This approach was widely viewed as a major mistake whose echoes we are still dealing with today. Economic, as opposed to regulatory, capital of the bank would not, as I have already noted, be changed by this special regulatory capital accounting, and such deductions from equity capital would not be reflected under GAAP.

Perhaps more to the point, it seems particularly relevant to underline the losses in financial markets-large losses can occur so quickly that regulators would be unable to close the failing operating sub as contemplated by H.R. 10 before the subsidiary's capital ran out. Indeed, losses might even continue to build, producing negative net worth in the subsidiary. At the time of closure of a subsidiary, there is nothing to prevent the total charges for losses against the parent bank's regulatory capital from exceeding the prior deduction required by H.R. 10. And closure and bankruptcy can and will be tied up in courts during which time the bank's capital and name are at risk. Our experience following the stock market crash of 1987-when a subsidiary of a major bank not only lost more than the bank's investment in its sub, but the bank was unable to dispose of the subsidiary for several years—underscores the seriousness of such concerns.

While contemplating movements in stock prices, let me note that merchant banking is potentially the most risky activity that would be authorized by H.R. 10, and would be especially risky for the insured bank if permitted to be conducted in bank subsidiaries.

Let me close, Mr. Chairman, by noting again that the Board is a strong advocate of financial modernization in order both to eliminate the inefficiencies of the current Great Depression regulatory structure and to create a system more in keeping with the technology and markets of the 21st century. We strongly support the thrust of H.R. 10 to accomplish these objectives. Equally as strongly, however, we also believe that the new activities should be authorized for banks through the holding company structure. That structure, especially for the new activities, also has the significant benefit of promoting effective supervision and the functional regulation of different activities. The holding company structure, along with the so-called "Fed-lite" provisions in H.R. 10, focuses on and enhances the functional regulation of securities firms, insurance companies, insured depository institutions and their affiliates by relying on the expertise and supervisory strengths of different functional regulators.

Thank you very much, Mr. Chairman. I request that my full remarks be included for the record.

[The prepared statement of Hon. Alan Greenspan follows:]

PREPARED STATEMENT OF HON. ALAN GREENSPAN, CHAIRMAN, BOARD OF
GOVERNORS, FEDERAL RESERVE SYSTEM

I would like to thank the Committee for the opportunity to present the views of the Federal Reserve on the current version of H.R. 10, the approach to financial modernization most recently approved by the House Banking Committee. Last year, I testified at length before this Committee on many of the issues related to your deliberations on this legislation. Our views have not changed on the need to mod

ernize our banking and financial system, on consolidated supervision, on the emphasis on reduced regulation, on the unitary thrift loophole, and especially on continuing to prohibit banks from conducting through their subsidiaries those activities that they are prohibited to do themselves. In the interest of time, however, I thought it might be best if I limit my formal comments only to the latter, that is, the setting of the underlying structure of American banking in the 21st century. The issue is whether the important new powers being contemplated are exercised in a financial services holding company through a non-bank affiliate or in a bank through its subsidiary. Such a decision would be of minor significance, and decidedly not a concern of legislators and regulators, if banks were not subsidized.

We at the Federal Reserve strongly support the new powers that would be authorized by H.R. 10. We believe that these powers, however, should be financed essentially in the competitive market place, and not financed by the sovereign credit of the United States. This requires that the new activities be permitted through holding companies and prohibited through banks.

Operating Subsidiaries

The Board believes that any version of financial modernization legislation that authorizes banks to conduct in their subsidiaries any activity as principal that is prohibited to the bank itself, is potentially a step backward to greater federal subsidization, and eventually to more regulation to contain the subsidies. I and my colleagues, accordingly, are firmly of the view that the long-term stability of U.S. financial markets and the interests of the American taxpayer would be better served by no financial modernization bill rather than one that allows the proposed new activities to be conducted by the bank, as proposed by H.R. 10. For reasons I shall discuss shortly, the Board is not dissuaded from this view by provisions that have been incorporated in H.R. 10 to address our concerns.

Subsidies. Government guarantees of the banking system-deposit insurance and direct access to the Fed's discount window and payments system guarantees-provide banks with a lower average cost of capital than would otherwise be the case. This subsidized cost of capital is achieved through lower market risk premiums on both insured and uninsured debt, and through lower capital than would be required by the market if there were no government guarantees. The lower cost of funding gives banks a distinct competitive advantage over nonbank financial competitors, and permits them to take greater risks than they could otherwise.

The safety net subsidy is reflected in lower equity capital ratios at banks, that are consistently below those of a variety of nonbank financial institutions. Importantly, this is true even when we compare bank and nonbank financial institutions with the same credit ratings: banks with the same credit ratings as their nonbank competitors are allowed by the market to have lower capital ratios. While the differences in capital ratios could reflect differences in overall asset quality, there is little to suggest that this factor accounts for more than a small part of the difference.

Under H.R. 10, the subsidy that the government provides to banks as a byproduct of the safety net would be directly transferable to their operating subsidiaries to finance powers not currently permissible to the bank or its subsidiaries. The funds a bank uses to invest in the equity of its subs are available to the bank at a lower cost than that of any other potential investor, save the United States Government, because of the subsidy. Thus, operating subsidiaries under H.R. 10 could conduct new securities, merchant banking, and other activities with a government subsidized competitive advantage over independent firms that conduct the same activity. That is to say, the use of the universal bank structure envisioned in H.R. 10 means the transference of the subsidy to a wider range of financial businesses, producing distortions in the competitive balance between those latter units that receive a subsidy and identical units that do not-whether those units are subs of holding companies or totally independent of banking.

H.R. 10 does not contain provisions that effectively curtail the transfer of the subsidy to operating subsidiaries or address this competitive imbalance. The provisions of H.R. 10 that would require the deduction of such investments from the regulatory capital of the bank (after which the bank must still meet the regulatory definition of well-capitalized) attempt, but fail, to limit the amount of subsidized funds that an individual bank can invest in its subs. What matters is not regulatory capital, but actual or economic capital. The vast majority of banks now hold significantly more capital than regulatory definitions of "well-capitalized" require. This capital is not "excess" in an economic sense that is somehow available for use outside the bank; it is the actual amount required by the market for the bank to conduct its own activities. The actual capital maintained by a bank is established in order to earn the perceived maximum risk-adjusted rate of return on equity. Unless this op

56-607 99-2

timum economic capital is equal to, or less than, regulatory capital, deductions from regulatory capital would in no way inhibit the transfer of the subsidy from the bank to the subsidiary.

Some have argued that the subsidy transference to subsidiaries of banks is no different from the transfer of subsidized bank dividends through the holding company parent to holding company affiliates. The direct upstreaming of dividends by a bank to its holding company parent that in turn invests the proceeds in subsidiaries of the holding company, while legally permissible, in fact does not occur and for good reasons, as I will explain below. In the 1990's, dividend flows from banks to their parent holding companies have been less than the sum of holding company dividends, interest on holding company debt, and the cost of holding company stock buy backs, a substitute for dividends. Thus, the empirical evidence indicates that, on net, at the largest organizations there has been no financing of a bank's holding company affiliates with subsidized equity of the associated banks. All of that part of the subsidy reflected in earnings has flowed to investors. (There are a few large individual institutions that have, in some years, upstreamed dividends in excess of investor payments, but the cumulative amounts are very small and the conclusions are unchanged.)

That bank dividends are not used to finance holding company subsidiaries should not be surprising. It simply is not in the interest of the consolidated banking organization to increase bank dividend flows beyond parent company capital-servicing cash flow needs because the resultant decline in bank capital would increase funding costs of the bank. Research at the Federal Reserve indicates that, over the past quarter of a century, for the largest banks the cost of uninsured bank funds has tended to rise as a bank's capital ratio fell and vice-versa. This is just what one should expect: As the risk-absorbing equity cushion falls, the risk for uninsured creditors rises. The flow of dividends from the bank to the parent holding company reduces bank capital. That reduction, in turn, reduces the risk buffer for uninsured creditors, increasing the funding cost of the bank on all the uninsured liabilities by more the data show-than the small subsidy transference of funding the additional equity investment in the affiliate.

Thus, were a bank holding company to finance its nonbank affiliates from bank dividends—that is, to directly pass on the bank's subsidy to the holding company's affiliates-the profitability of the consolidated organization would decline. If there were no net costs to the bank from upstreaming dividends to its parent for affiliate funding, it would be the prevalent practice today. In short, the subsidy appears to have been effectively bottled up in the bank. The Federal Reserve Board believes that this genie would be irreversibly let out of the bottle, however, should the Congress authorize wider financial activities in operating subs. Subsidized equity investments by banks can be made in their own subsidiaries without increasing funding costs on all of the bank's uninsured liabilities because the consolidated capital of the bank would not change in the process. But since the activities authorized to banks' subsidiaries cannot differ from those available to the bank itself, there is no additional profit to the overall banking organization in shifting bank powers to a subsidiary.

But H.R. 10 would permit activities not now permitted in a bank. Those activities, when performed in bank subsidiaries and financed with bank equity capital would increase the potential profit to the overall banking organization. It would also inevitably induce the gravitation to subsidiaries of banks, not only of the new powers authorized by H.R. 10, but all of those powers currently financed in holding company affiliates at higher costs of capital than those available to the bank. H.R. 10 thus effectively authorizes all holding company powers to be funded in the bank at funding costs significantly lower than the funding costs of its holding company.

For the 35 of the 50 largest bank holding companies for which comparisons are available, ratings on debentures are always somewhat higher at the bank than at the holding company parent and, of course, higher ratings translate into lower interest rates. As might be expected, the data show that the value of these differences in bond ratings is higher during periods of market stress, when subsidies are more valuable, because the market is more risk sensitive. But even today, when losses in the financial system are quite low, the cost of debt capital to banks still averages 10 to 12 basis points below that of the parent holding companies. That difference in bond ratings today between banks and bank holding companies, let alone the larger difference between banks and other financial institutions, is a significant part of the 20 to 30 basis point gross margin on A-rated or better investment grade business loans more than enough significantly to change lending behavior if it were not available.

Business loan markets are particularly competitive, and hence there is little leeway for a competitor with higher funding costs to pass on such costs to the bor