Why We Need an "Accord" for Federal Reserve Credit Policy: A Note.

Goodfriend, Marvin

The 1951 Accord between the Treasury and the Federal Reserve was one of the most dramatic events in U.S. financial history. The agreement liberated monetary policy from the commitment, dating from World War II, to support government bond prices. It reasserted the principle of Federal Reserve independence so that monetary policy might serve primarily as an instrument for macroeconomic stabilization.

The Federal Reserve, however, executes both monetary and credit policies, and no Accord has yet been established for its credit policies. The reason is that, until recently, fiscal concerns have not threatened the misuse of Fed credit policies in the way that bond price supports did for monetary policy. Large federal budget deficits, a deposit insurance crisis, or significant foreign exchange market intervention could change that. [1] Just as the 1951 Accord greatly improved monetary policy, an Accord for Fed credit policy established today, while fiscal concerns are still relatively small, could yield significant benefits in the future.

1. MONETARY VERSUS CREDIT POLICY

Distinguishing between monetary and credit policy is straightforward. [2] Monetary policy refers to changes in the stock of high-powered money, that is, currency plus bank reserves, accomplished by open market operations in domestic securities or foreign exchange. For example, a central bank takes a monetary policy action if it increases bank reserves by purchasing securities. Credit policy, on the other hand, changes a central bank's assets while holding the stock of high-powered money fixed. For example, a central bank takes a credit policy action when it uses funds obtained by selling Treasury securities to acquire other assets. Credit policies also include regulation and supervision of the banking system, but such aspects of policy will not be discussed here.

2. THE ACCORD PRINCIPLES FOR CREDIT POLICY

The 1951 Accord established the principle that monetary policy should be used to stabilize the macroeconomy, regardless of the fiscal concerns of the Treasury. It restored the idea that a fully independent central bank contributes importantly to economic stability. [3] Independence insulates the Fed from short-run inflationary pressures to stimulate employment and help finance the Treasury. It also frees the Fed from having to get Congressional or Treasury approval for its policy actions, enabling the Fed to react quickly to short-run macroeconomic or liquidity shocks.

Congress bestows such independence only because it is necessary for the central bank to do its job effectively. Hence, the presumption ought to be that the Fed should perform only those functions that must be carried out by an independent central bank. Monetary policy is both necessary and sufficient to pursue macroeconomic stabilization policy and to deter system-wide liquidity crises. Credit policy directs funds promptly to illiquid institutions when macroeconomic conditions do not call for a change in high-powered money.

This suggests the following Accord principles for Fed credit policy: (1) liquidity assistance should not fund insolvent institutions; (2) credit policy should not fund expenditures that ought to get explicit Congressional authorization; (3) Congress should not direct the Fed to transfer assets to the Treasury in order to reduce the Federal deficit.

Three Fed credit policies discussed below illustrate the above concerns. First, liquidity assistance potentially provides funds to insolvent institutions and raises the cost of deposit insurance. Second, Fed credit policy may inappropriately finance sterilized foreign exchange market intervention and some foreign expenditures of the Treasury. Third, the transfer of Fed surplus assets to the Treasury, as directed by Congress, potentially weakens Fed independence. In each case, an Accord for Fed credit policy would help implement the above principles.

3. LIQUIDITY ASSISTANCE

As a rule, the Fed finances liquidity assistance to depository institutions with funds acquired by selling Treasury securities--leaving high-powered money unchanged. Thus, as mentioned above, liquidity assistance is a credit policy. In practice, the Fed fully collateralizes its discount-window lending. Its supervisory role enables it to value bank loans for purposes of collateral prior to any request for funds. Moreover, the Fed can lend on less than full assessed collateral value to further protect itself. Hence, discount-window lending has involved little risk for the Fed.

Discount-window credit can save a temporarily illiquid but solvent bank. But discount-window loans potentially allow a truly insolvent bank, by pledging collateral to the discount window, to more easily pay out uninsured depositors prior to being closed. Such lending imposes costs on the deposit insurer, when it delays a declaration of insolvency, by moving uninsured depositors from last to first in line.

Because Fed liquidity assistance must be extended promptly, it is impractical for Congress to authorize each provision. Without Congressional guidance, however, Fed lending may not take into account potential losses it might impose on the deposit insurance fund, or the taxpayer, if an illiquid bank to which it is lending turns out to be insolvent. Lending on acceptable collateral is safe from the Fed's point of view, but, as mentioned above, there are times when it may delay the closing of an insolvent bank by paying out uninsured depositors at the expense of the deposit insurance fund or the taxpayer. [4]

The 1991 Federal Deposit Insurance Corporation Improvement Act (FDICIA) recognized the need for a mechanism to encourage the Fed to withdraw its credit line soon enough to protect the insurer and the taxpayer. FDICIA provides incentives for the Fed not to lend to undercapitalized banks. [5] To the extent that capitalization continues to be measured largely on book rather than market valuation, however, there may be instances when the new law is less than fully effective.

An Accord could be arranged (with Congressional help) between the Fed, the Treasury, the deposit insurers, and the depository institution chartering agencies to better ensure that liquidity assistance does not delay the closure of insolvent banks. One possibility would be to have the Fed stop lending when, on its estimate of market values, a liquidity problem is judged to become a solvency problem. A second option would be to agree on a rule limiting the share of assets that a bank might pledge to the Fed. This would mimic the "negative pledge clauses" in private bond covenants designed to protect bond holders against asset stripping by managers in the run-up to bankruptcy. Of course, if it seems feasible and desirable, an Accord could involve more elaborate coordination.

4. STERILIZED FOREIGN EXCHANGE MARKET INTERVENTION AND WAREHOUSING

Two agencies conduct official foreign exchange market intervention in the United States--the Treasury, through its Exchange Stabilization Fund (ESF), and the Federal Reserve, under the guidance of the Federal Open Market Committee--with intervention coordinated between the two. As a mechanical matter, intervention is simply a purchase of foreign currency, with U.S. dollars, in the foreign exchange market.

A Fed purchase of foreign exchange that increases high-powered money is monetary policy, but an acquisition of foreign exchange funded by selling dollar-denominated securities is credit policy. The latter is commonly known as sterilized foreign exchange intervention because its potential effect on high-powered money is offset by the sale of securities. The Fed undertakes sterilized intervention for its own account and for the ESF. Such intervention is sometimes undertaken in cooperation with foreign monetary authorities using reciprocal currency arrangements. These are, in effect, lines of credit giving central banks access to each other's currency. [6]

The ESF borrows dollars to buy foreign exchange by using its foreign exchange purchases as RP collateral at the Fed--a practice known as foreign exchange warehousing. [7] In effect, the ESF finances its foreign exchange portfolio much as, say, dealers use RPs to finance their security portfolios. The Fed routinely sterilizes the effect on high-powered money of its dollar-denominated lending to the ESF by selling an equivalent value of dollar-denominated securities. Whether or not sterilized foreign exchange intervention is carried out by the Fed for its own account or for the ESF, the net result is to substitute foreign-currency-denominated securities (or interest-earning deposits at a foreign central bank) for dollar-denominated securities on the Fed's balance sheet, without changing high-powered money.

There is little evidence that large-scale sterilized intervention has a sustained effect on the exchange rate. [8] In some situations, sterilized intervention may temporarily stabilize the exchange rate; or it may signal government resolve to follow up with monetary or fiscal policy actions that will powerfully influence the exchange rate in the future. To the extent that such intervention needs to be carried out promptly, without public debate, it may be useful for an independent central bank to finance it. Nevertheless, in light of the ineffectiveness of sterilized intervention, Congress could explicitly limit the use of Fed credit policy for this purpose. Of course, the Fed and the Treasury could agree to keep sterilized intervention to a minimum in lieu of Congressional action.

Foreign Exchange Warehousing

In conjunction with the proposed limit on sterilized foreign exchange intervention, an end to warehousing would further implement the second Accord principle. The ESF has occasionally made loans, by short-term swap agreements and by other means, to heavily indebted countries for balance of payments purposes and to help manage their external debt. [9]

The ESF could clearly carry out such responsibilities without the help of the Fed. If need be, the ESF could be provided with additional funds borrowed by the Treasury itself, or the ESF could be given additional authority by Congress to borrow on its own account.

When the ESF finances itself by warehousing foreign exchange with the Fed, a sale of Treasury securities to the public is also the ultimate source of funds. The only difference is that the Treasury securities are not newly issued, but rather sold from the Fed's portfolio. It is, however, as if the debt were newly issued, since the Fed simply returns to the Treasury the interest it receives on the Treasury securities it holds.

The main difference between Fed financing, and financing by the Treasury itself, is that the former is arranged between Fed and Treasury officials without an explicit appropriation from Congress. A second difference is that Fed financing does not show up as a measured increase in the Federal deficit, since it does not involve newly issued debt.

Whatever financing method is adopted, loans made to help foreign governments finance their balance of payments deficits or to manage their external debt are clearly deficit-financed fiscal policy actions of the U.S. government. As is the case with any fiscal policy, the presumption is that Congress should authorize the spending and explicitly appropriate the necessary funds. Since Fed warehousing for the Treasury does not require Congressional authorization and obscures the funding, warehousing would not appear to be an appropriate use of Fed credit policy.

5. THE TRANSFER OF FED SURPLUS TO THE TREASURY

The Deficit Reduction Act passed by the U.S. Congress in 1993 contains a provision to take $213 million from the Fed's surplus account to help meet budget reconciliation targets in 1997 and 1998.10 Surplus is a capital account on the Fed's balance sheet, a kind of retained earnings for contingencies. The transfer of surplus is tiny when compared to total Fed assets, which were approximately $370 billion at the end of 1992, about $330 billion of which were security holdings. In fact, the transfer is only about 7 percent of the Fed's $3 billion end-of-1992 surplus.

Although it is small, the transfer is important because it represents a kind of policy action that, if resorted to routinely in the future, could eventually shrink the volume of liquid assets in the Fed's portfolio enough to undermine the central bank's monetary and credit policy powers, and ultimately, its financial and political independence as well. Moreover, as we shall see below, although the transfer of Fed assets appears to provide supplementary funds to the Treasury, in fact, it provides no additional revenue. For these reasons, Congress should agree to an Accord not to transfer Fed surplus to the Treasury.

Historical Precedent for the Transfer of Fed Surplus

The Federal Reserve Act authorized the Fed to build up a surplus by retaining interest earned from its asset portfolio until surplus reached 40 percent of paid-in capital of member banks." In 1919 the law was changed to allow surplus to be raised to 100 percent of subscribed capital (twice paid-in capital). In 1933, half of Fed surplus, $139 million, was used by Congress to capitalize the newly established Federal Deposit Insurance Corporation.

The 1959 Federal deficit of $13 billion was three times larger than any previous peacetime deficit and the next five years saw a string of deficits that generated Congressional pressure for the Fed to cut its surplus. In 1964 the Fed announced a voluntary reduction of surplus, reducing it to paid-in capital. That decision added $524 million to the amount that the Fed paid to the Treasury in 1965. The Fed has held surplus equal to paid-in capital since then. As a result of the new legislation, surplus will be kept equal to paid-in capital minus $213 million.

Budget Mechanics of the Transfer of Fed Surplus to the Treasury

The Fed will obtain the funds to make the required transfer by selling Treasury securities from its portfolio to the private sector. The Treasury will receive the $213 million as additional revenue in 1997-98, and thus record a smaller deficit for those years.

As long as the Treasury uses the supplementary revenue to cut back on borrowing or to finance additional spending, the transfer will not affect the stock of high-powered money in the hands of banks and the public. Hence, the transfer is not a monetary policy action. Rather it's a credit policy action that can be thought of as an interest-free loan from the Fed to the Treasury financed by a sale of securities from the Fed's portfolio, reflected in a shrinking of the Fed's capital account.

The transfer of assets to the Treasury is intended to provide it with a onetime supplemental source of funds to help narrow the Federal deficit. To see that it will not in fact do so, consider the Treasury securities the Fed will sell to get the $213 million for the transfer. When the Fed holds these securities, it is as if they are extinguished from the Treasury's point of view, because the Treasury pays the interest to the Fed and the Fed simply returns that interest to the Treasury. Once the Fed sells the securities to the public, however, the Treasury no longer gets back its interest payments.

In short, selling securities from the Fed's surplus account and transferring the proceeds to the Treasury is equivalent to the Treasury issuing new debt to borrow the funds directly from the public. The transfer of Fed surplus will have no effect on the correctly measured Federal deficit. The transfer of Fed assets to the Treasury will merely appear to reduce the Federal deficit because the sale of securities held by the Fed is not recorded as a new issue of Treasury debt.

The Role of Fed Surplus and Federal Reserve Independence

Surplus is employed in commercial enterprises as a reserve for contingencies such as absorbing losses or meeting expenses and dividends when earnings are low. The Fed employs its surplus in a similar manner. The most important contingencies are exchange rate revaluation of foreign-currency-denominated securities that the Fed holds for its own account. Since the Fed marks these assets to market monthly, an appreciation of the foreign exchange value of the dollar reduces the dollar value of the Fed's foreign-security holdings. The Fed carries its dollar-denominated securities at historical cost. But surplus is also used to absorb any realized losses on sales of domestic securities.

Currently, the Fed pays its interest earnings to the Treasury weekly. Starting from zero, the Fed accrues payments each week as so-called undistributed net income and turns it over to the Treasury with a week lag. In 1992, for example, net interest earnings averaged around $325 million a week, and at the end of the year the Fed held about $22 billion of foreign-currency-denominated securities. [12] Although not all of the $22 billion was held for the Fed's own account, the magnitudes are such that a monthly appreciation of the dollar on the foreign exchange market could significantly offset net interest income in a given week.

As an accounting matter, undistributed net income is not allowed to go negative. Whenever a revaluation of foreign security holdings or a realized loss on the domestic portfolio causes it to do so, assets are moved from the surplus account to bring undistributed net income back up to zero. In the following weeks, no transfers are made to the Treasury until the Fed's assets are replenished and surplus is restored to the level of paid-in capital. In general, any gains or losses on foreign securities that the Fed holds for its own account show up as larger or smaller Fed payments to the Treasury. Profits or losses on warehoused foreign securities accrue to the ESF.

Surplus, then, serves as a buffer helping to protect paid-in capital and to insure that the Fed's liquid securities cover its high-powered money liabilities. Eliminating even the entire $3 billion surplus account would only reduce the Fed's portfolio of securities by about 1 percent, so it would certainly not impair the Fed's ability to conduct policy. The risk is that the elimination of surplus would undermine the principle that the Fed should retain possession of the interest earning assets it acquires through money creation. That might tempt Congress to order even more transfers in the future.

If carried far enough, stripping the Fed of its liquid assets would obviously interfere with its ability to conduct monetary and credit policy. Equally important, however, it would undermine the Fed's financial independence by denying it enough interest income to finance its operations without having to ask Congress for appropriations or resorting to inflationary money creation. The excess of Fed earnings over expenses has been large recently--the Fed paid about $17 billion to the Treasury in 1992. [13] But excess earnings could be reduced in the future if nominal interest rates come down, reserve requirements are reduced further, or interest is paid on required reserves. Meanwhile, the excess is simply returned to the Treasury.

Thus, surplus serves as a bulwark protecting both the financial independence of the Fed and its monetary and credit policy powers. Moreover, the Fed's financial independence is the foundation of its political independence, so respect for Fed surplus on the part of Congress would strengthen the Fed's determination to pursue noninflationary monetary policy.

6. CONCLUSION

The Federal Reserve pursues both monetary and credit policies. Yet no Accord protects its credit policies from fiscal misuse the way the 1951 Accord protects monetary policy. With that in mind, the paper presented some principles for credit policy, and proposed Accords that would implement those principles for three prominent policies. The basic idea is that Congress has provided the Fed with the independence necessary to carry out central bank functions effectively, and the Fed should perform only those functions.

In effect, FDICIA already partially incorporates an Accord to limit the cost that liquidity assistance potentially imposes on the deposit insurance fund. That Accord may have to be strengthened, however, to more effectively restrict liquidity assistance to institutions that have become insolvent on a market value basis.

Since there is little evidence that sterilized foreign exchange intervention has more than a temporary effect on the exchange rate, the Fed and the Treasury could reach an Accord to keep such intervention to a minimum. Foreign exchange warehousing could also be ended by a simple agreement between the Fed and the Treasury. But Congress could explicitly limit the potential abuse that warehousing exemplifies: the use of Fed credit policy for off-budget funding without explicit Congressional authorization.

The last policy considered was the transfer of Fed surplus to the Treasury. This credit policy has budget consequences in appearance only. Nevertheless, it could set a harmful precedent for further stripping the Fed of assets that would ultimately weaken the central bank's independence and its ability to conduct policy.

The author is Senior Vice President and Policy Advisor at the Federal Reserve Bank of Richmond. This article originally appeared in the Journal of Money, Credit, and Banking, vol. 26 (August 1994), and was prepared for the October 1993 Federal Reserve Bank of Cleveland/Journal of Money, Credit, and Banking Conference "Federal Credit Allocation: Theory, Evidence and History." The views are the author's and do not necessarily reflect those of the Federal Reserve Bank of Richmond or the Federal Reserve System.

(1.) Fry (1993) reviews the fiscal activities that governments in a sample of twenty-six developing countries order their central banks to undertake.

(2.) This distinction is used extensively by Goodfriend and King (1988).

(3.) Stein (1969) contains an excellent discussion of events leading up to the 1951 Accord.

(4.) Schwartz (1992) discusses numerous examples of discount-window lending to insolvent institutions. Garcia (1990) catalogs some nontraditional uses of the discount window.

(5.) See The Federal Reserve Discount Window, 1994.

(6.) Fisher (1994), p. 4, lists the Federal Reserve's current reciprocal currency arrangements.

(7.) See Crain (1990). The ESF also finances itself by other means, see Exchange Stabilization Fund Annual Reports.

(8.) See, for example, Bordo and Schwartz (1990), Edison (1992), and Obstfeld (1988).

(9.) See the "operations statements" in Exchange Stabilization Fund Annual Reports.

(10.) See the Omnibus Budget Reconciliation Act of 1993.

(11.) The historical treatment of surplus is discussed in Goodfriend and Hargraves (1983), together with the history of Fed payments to the Treasury.

(12.) Board of Governors of the Federal Reserve System 1992 Annual Report, p. 262. The combined foreign exchange holdings of the Federal Reserve and the Treasury nearly reached $45 billion in December 1989 (Jacobson 1990).

(13.) Board of Governors of the Federal Reserve System 1992 Annual Report, pp. 276-77.

REFERENCES

Board of Governors of the Federal Reserve System. 1992 Annual Report. Washington, D.C.: Board of Governors of the Federal Reserve System.

Bordo, M., and A. Schwartz. "What Has Foreign Exchange Market Intervention since the Plaza Agreement Accomplished?" Working paper no. 3568. Cambridge, Mass.: National Bureau of Economic Research, December 1990.

Crain, B. "Warehousing and the Exchange Stabilization Fund." U.S. Congress, House Banking Committee, Subcommittee on Domestic Monetary Policy, August 1990.

Edison, H. "The Effectiveness of Central Bank Foreign Exchange Intervention: A Survey of the Post-1982 Literature." Washington, D.C.: Board of Governors of the Federal Reserve System, June 1992.

Exchange Stabilization Fund Annual Reports, U.S. Department of the Treasury, Washington, D.C.

The Federal Reserve Discount Window. The Federal Reserve System, revised, 1994.

Fisher, P. "Treasury and Federal Reserve Foreign Exchange Operations." Federal Reserve Bulletin, January 1994, 1-4.

Fry, M. "The Fiscal Abuse of Central Banks." International Monetary Fund Working paper. Washington, D.C.: International Monetary Fund, July 1993.

Garcia, G. "The Lender of Last Resort: Recent Developments and Nontraditional Examples." U.S. Senate, Committee on Banking, Housing and Urban Affairs, December 1990.

Goodfriend, M. "Money, Credit, Banking, and Payments System Policy." In The Payment System: Risk, Efficiency, and the Role of the Federal Reserve, edited by D. Humphrey, pp. 247-77. Boston: Kiuwer, 1990. Reprinted in Federal Reserve Bank of Richmond Economic Review (January/February 1991), 7-23.

Goodfriend, M., and M. Hargraves. "A Historical Assessment of the Rationales and Functions of Reserve Requirements." Federal Reserve Bank of Richmond Economic Review (March/April 1983), 3-21.

Goodfriend, M., and B. King. "Financial Deregulation, Monetary Policy and Central Banking." In Restructuring Banking and Financial Services in America, edited by W. Haraf and R. Kushmeider, pp. 216-53. Boston: Kluwer, 1990. Also in Federal Reserve Bank of Richmond Economic Review (May/June 1988), 3-22.

Jacobson, K. "U.S. Foreign Exchange Operations." Federal Reserve Bank of Kansas City Economic Review (September/October 1990), 37-50.

Obstfeld, M. "The Effectiveness of Foreign Exchange Intervention: Recent Experience." Working paper no. 2796. Cambridge, Mass.: National Bureau of Economic Research, December 1988.

Omnibus Budget Reconciliation Act of 1993, Public Law 103-66, U.S. Congress, 103rd Session Public Laws, 1993.

Schwartz, A. "The Misuse of the Fed's Discount Window." Federal Reserve Bank of St. Louis Economic Review (September/October 1992), 58-69.

Stein, H. "The Liberation of Monetary Policy." In The Fiscal Revolution in America, pp. 241-80. Chicago: University of Chicago Press, 1969.