A History of the Federal Reserve, Volume 2 (12 page)

Martin won the skirmish with Sproul over bills-only. He had the better argument: sporadic intervention in the long-term government securities market would lead market participants to anticipate support of particular issues during Treasury sales or market volatility. The Federal Reserve should intervene only for general monetary policy operations by supplying or withdrawing reserves. A “free market” would not develop if the Federal Reserve supported long-term bonds. Characteristically, Martin rejected the statistical tests that Sproul’s staff and others developed to show that the long-term market had not developed greater depth, breadth, and resiliency. He did not deny that intervention in the long-term market would change long-term rates, an important issue for Sproul. He conceded that effects on long-term rates might be slower under bills-only. He wanted to prevent the market from depending on, and waiting for, System intervention. Sproul and his staff would not acknowledge the importance of this argument (memo, Gaines to Sproul, FOMC Correspondence, April 27, 1956).

Long-term rates move around an average (geometric mean) of expected future short-term rates. Interventions to change a long-term rate must also change other prices up and down the maturity structure. Later, research developed the expectations theory of the term structure, showing why interventions of the kind Sproul proposed were unlikely to succeed for long if at all.
²⁰

Sproul thought the FOMC was “foolish” to agree “for the record that we enter into open market operations solely to supply or absorb reserves” (letter, Sproul to Powell, Sproul papers, Correspondence 1952–55, September 17, 1954, 2). He believed that “this business of central banking requires a little more complicated approach than just ‘putting in and taking out reserves’ so as to maintain some figure or range of ‘free reserves.’
The cost of credit at short- and long-term is a central banking consideration as well as the supply of bank reserves” (ibid., 1). He believed also that “aid can properly and appropriately be extended to Treasury refunding operations by transactions in rights, when issued securities, and securities of comparable maturity; that it can be extended at times, more effectively in this way than by confining our operations to Treasury bills” (letter, Sproul to Martin, Sproul
papers, Correspondence 1952–55, December 4, 1953, 4). He doubted that “the market could develop into a broad, impersonal mechanism in which gradual and orderly changes in demand and supply conditions would be reflected in equally orderly changes in prices and yields, if only the System Account were out of the picture so that dealers might rationalize their expectations. The minority [New York] . . . viewed the market as one in which the major disturbing forces were the so-called ‘natural forces’ not the System Account. The principal uncertainty created by the Federal Reserve System stems from the possibility of shifts in its basic policy not from the areas of the market in which it might operate”
²¹
(Sproul papers, Q and A for Flanders hearings, questions 3, November 1954).

20. Modern work on the term structure did not begin until Meiselman (1962). By that time, bills-only had ended. Until the financial disturbances in 2008, the Federal Reserve held few long-term bonds. In 2008, more than half its portfolio was longer term.

For Sproul’s argument in favor of intervention to be correct, the desk would have to distinguish correctly between temporary and permanent changes in prices at different maturities. As the market learned to anticipate intervention, prices and yields would change, as Martin insisted. Sproul never tried to show that the manager would know how much to intervene and when to do so. He assumed that intervention would stabilize the market. Time and experience have not supported Sproul’s views that the Federal Reserve could change the relative prices of securities of different durations other than temporarily or change security prices independently of changes in the stock of reserves or base money. Most central banks operate mainly or exclusively in the short-term market although they may buy and hold long-term securities. In the United States, Congress put pressure on the Federal Reserve at times to buy agency securities to encourage housing.

At the Board, Riefler put considerable weight on the role of forwardlooking anticipations, long before they received much attention in academic work. He described government security dealers and market professionals as always trying “to ascertain the significance of all System policy actions” and adjusting their portfolios accordingly (Riefler, 1958a, 5). Markets, he said, do not try to counter System operations because of the size of the System’s portfolio and because the System is a central bank, concerned about its policy objectives not its profitability.

Martin did not insist that policy could do no more than change the stock of reserves and let the market determine the relative prices of different
maturities. Theoretical arguments of this kind had little appeal for him. He would have dismissed the claim that he had adopted a “monetarist” position that the stock of reserves and base money were the key variables. He based his decision on the “practical” argument that, if the Federal Reserve intervened to support particular issues, it would soon be back pegging interest rates for the Treasury. He added, moreover, that there would never be a “free market” in Treasury securities if the market expected the Federal Reserve to limit price changes by intervening in the long-term market.
²²

21. Even if one accepted that the principal source of market disturbance is Federal Reserve intervention, additional intervention or uncertainty about where the Federal Reserve would intervene would not smooth the market. New York refused to accept the Board’s argument about dealer expectations and the suggestion that intervention would be a step toward pegging. To New York’s chagrin, market newsletters emphasized this point (Gaines to Rouse, Sproul papers, bills-only, June–October 1954, June 9, 1954).

The weak point in Martin’s argument and the staff’s support for bills-only was a failure to recognize that support for Treasury operations, whether by purchasing bills or bonds, increased the risk of inflation. Some of the staff, and possibly Martin at times, recognized the need to control money growth if they were to prevent inflation. They did not acknowledge that support of Treasury financings made monetary control more difficult. This became a problem in the 1960s.

Martin won the System’s economic and policy dispute but lost the broader political skirmish.
²³
Most academic economists who expressed their views on bills-only sided with Sproul. In journal articles and congres
sional hearings, they criticized bills-only as an unnecessary and costly restriction. Congressional opposition to bills-only gained strength as interest rates rose during the decade. The issue united Senator Douglas and Congressman Patman, who disagreed totally about pegged rates but agreed for different reasons that bills-only was wrong. In 1960, twenty-one senators urged the System to end bills-only and increase money growth to equal the rate of output growth (Board Minutes, March 15, 1960).
²⁴
The Federal Reserve’s efforts to defend the policy attracted few supporters.

22. In a 1954 hearing, Martin responded to a written question by highlighting the uncertainty created by sporadic intervention borrowing in part from his speech on free markets. “The constant possibility of official action, which from the standpoint of investors and market intermediaries would often seem capricious, constituted a market risk which private investors could in no reasonable way anticipate and evaluate” (Joint Committee on the Economic Report, 1954, 24). Testifying to the Joint Economic Committee in 1959, Martin listed three ways that open market operations affect the economy. “(1) They change the volume of reserves. . . . (2) They affect the volume of securities . . . (3) They influence the expectations of professional traders and investors regarding market trends” (Martin, 1959, 19). The reference to expectations reflects Riefler’s influence and gives an emphasis that did not appear in the academic literature for many years. Martin put the most weight on the change in reserves. This emphasis is also in other Federal Reserve publications at this time. Within the Federal Reserve, Sproul had been the leading opponent of pegged rates and neither wanted nor intended to go back to pegging. He seemed unable to accept that the long-term market would function efficiently without support. Martin, for his part, did not address the issue that bothered Sproul: if the Federal Reserve intervened in the short-term market, the market was not “free” but depended on the System’s intervention. Martin’s claim was that the Federal Reserve could not know the proper relationship between prices of different maturities (Joint Committee on the Economic Report, 1954, 235). Sproul did not respond directly to this argument. Martin also claimed that, unlike Sproul, he believed arbitrage occurred quickly (ibid., 230).

23. An example of press commentary is an editorial in the
Journal
of
Commerce,
a business newspaper. The December 14, 1954, editorial blamed bills-only for the declining share of long-term bonds in the Treasury market. The editorial made no mention of the Treasury’s unwillingness to pay the higher interest rate on long-term bonds. The bills-only controversy illustrates the changed positions of New York and Washington. In the 1920s, Governor Benjamin Strong (New York) dominated policy actions and procedures. In the 1950s, New York could protest, but Washington decided.

Faced with intense, growing criticism of the bills-only policy, early in 1958 Chairman Martin appointed a special committee to review procedures after five years of experience. President Hayes initiated the review in a memo questioning whether the procedures and policies then in effect were optimal ways of achieving the System’s objectives while aiding Treasury finance and improving the functioning of the dealer market for government securities. Although the Board’s staff and the New York bank recognized some merit in each other’s position, and they agreed on objectives, basic differences remained.

The broad issue between the Board and the bank concerned how Federal Reserve operations should change to recognize the existence of a large, outstanding stock of debt, often growing and frequently refunded. Both accepted an obligation to assist the Treasury to an extent consistent with the System’s objectives. Growth of new, non-bank intermediaries was a second major change since the 1920s. Analysis at the time suggested that the existence of these intermediaries made a significant change in the transmission of monetary policy (see Gurley and Shaw, 1960).
²⁵
Both staffs’ responses accepted that markets would reflect the new conditions, but open market operations continued to work by changing the supply of reserves.

Much of the discussion restated earlier positions. The Board’s staff claimed that there were no occasions since 1953 “when operations in securities other than bills would have made it easier to accomplish monetary policy objectives” (Memo, “Analysis of Issues to be Considered by Special Committee,” Board Records, January 14, 1958, 3). The manager responded by citing some occasions when long- and short-term rates moved in op
posite directions, periods when he would have intervened in the long-term market.
²⁶
This did not respond directly to the Board’s statement, and it did not establish that the intervention would have been effective. In fact, aggressive reduction of short-term rates in a recession can increase expected future inflation, raising long-term rates.

24. Martin responded on behalf of the Board, pointing out that bills-only was not a rigid rule. The System had intervened in November 1955 and July 1958 and on two other occasions had exchanged its holdings for other than bills (August 1959 and February 1960) (Board Minutes, April 14, 1960, letter, Martin to 21 senators).

25. Some secondary issues arose also. How could financing of the dealer market be improved? The manager wanted opportunity to swap securities with market participants, an issue that the FOMC had considered and rejected earlier.

The Board’s staff and New York agreed that they could not produce evidence on the effectiveness of the policy by citing evidence from the longterm market. This required evidence on the counterfactual that no one could have. The Board’s staff argued that every time the market seemed on the verge of becoming disorderly, it “fairly quickly made its own adjustments” (“Analysis of Issues,” Board Records, January 14, 1958, 10). Support operations, they said, would have damaged the market’s ability to adjust. The staff added several administrative complications that would have arisen if the System had operated, at times, in the long-term market (ibid., 11).

The Board’s staff dismissed the suggestion that the FOMC should work to prevent, instead of correct, disorderly conditions. Prevention required forecasting and could quickly become an excuse for intervening more frequently, thereby increasing uncertainty for market participants. The staff then made the case for rules to reduce discretion: “Decision-making, moreover, is facilitated by having a set of recognized, well-conceived, generally workable ground rules and working principles as a basis for operations. . . . Perhaps more important is the desirability of public recognition of the principles, rules, and guides that govern monetary policies and operations”
²⁷
(ibid., 26).

The staff response also contained a general statement of “working principles” that explained the staff’s understanding of the way monetary policy affects the economy. The first principle is that “regardless of the instruments through which Federal Reserve credit is extended, the supply of bank reserves is affected” (ibid., 3). The report then discussed availability and the cost of credit as reflected in the correlation between free reserves and market interest rates, relative yields on securities, and the role of the government securities market. Missing is the earlier discussion of anticipations and the transmission of the monetary impulse via relative prices and real yields to the markets for output, investment, and consumer spending.

26. The manager presumed that the market was wrong and warranted correction. See his appendix to “The Ad Hoc Report—After Five Years” (Board Records, January 15, 1958). Differences in temporary and persistent changes in conditions would explain such movements.

27. This is a long, first step in the evolution toward greater transparency and what economists later called credibility and away from historic central bank secrecy. The staff, at this time, could see the advantages of a rule for intervention procedures but not for policy actions more generally.