President, Federal Reserve Bank of Richmond
The CATO Institute 29th Annual Monetary Conference
The financial crisis of 2007 and 2008 was a watershed event for the Federal Reserve and other central banks. The extraordinary actions they took have been described, alternatively, as a natural extension of monetary policy to extreme circumstances, or as a problematic exercise in credit allocation. I have expressed my view elsewhere that much of the Fed's response to the crisis falls in the latter category rather than the former.1 Rather than reargue that case, I want to take this opportunity to reflect on some of the institutional reasons behind the prevailing propensity of many modern central banks to intervene in credit markets. As always, these remarks are my own and the views expressed are not necessarily shared by my colleagues on the Federal Open Market Committee.2
There is widespread agreement among economists that a vigorous monetary policy response can be necessary at times to prevent a contraction from becoming a deflationary spiral. Financial market turmoil often sparks a flight to monetary assets. In the 19th and 20th centuries, this often took the form of shifts out of deposits and into notes and specie. Under a fractional reserve banking system, this necessitates a deflationary contraction in the overall money supply unless offset through clearinghouse or central bank expansion of the note supply. In modern financial panics, banks often seek to hoard reserve balances, which again would be contractionary absent an accommodating increase in the central bank reserve supply. In both cases, the need is for an increase in outstanding central bank monetary liabilities.
The Fed's response during the financial crisis was not purely monetary, however. In the first phase of the crisis — from the fall of 2007 through the summer of 2008 — its credit actions were sterilized; lending through the Term Auction Facility and in support of the merger of Bear Stearns and JPMorgan Chase was offset by sales of U.S. Treasury securities from the Fed's portfolio. (Note that such sterilized actions are equivalent to issuing new U.S. Treasury debt to the public and using the proceeds to fund the lending.) It wasn't until September 2008 that the supply of excess reserves began to increase significantly. This expansion was accomplished through the acquisition of an expanding set of private assets — loans to banks and other financial institutions and later mortgage-backed securities and debt issued by Fannie Mae and Freddie Mac. While some observers describe this phase of the Fed's response as a standard, monetary expansion in the face of a deflationary threat, the Fed's own characterization often emphasized instead the intent to provide direct assistance to dysfunctional segments of the credit markets.
Clearly, an equivalent expansion of reserve supply could have been achieved via purchases of U.S. Treasury securities — that is, without credit allocation.
Like the Fed, the European Central Bank and other central banks have pursued credit allocation in response to the crisis.
The impulse to reallocate credit certainly reflects an earnest desire to fix perceived credit market problems that seem within the central bank's power to fix. My sense is that Federal Reserve credit policy was motivated by a sincere belief that central banks have a civic duty to alleviate significant ex post inefficiencies in credit markets. But credit allocation can redirect resources from taxpayers to financial market investors and, over time, can expand moral hazard and distort the allocation of capital. This implies a difficult and contentious cost-benefit calculation. But no matter how the net benefits are assessed, central bank intervention in credit markets will have distributional consequences.
Central bank credit allocation is therefore bound to be controversial. Indeed, such actions taken by the Fed over the last few years have generated a level of invective that has not been seen in a very long time. Critics have sought to exploit the resentment of credit market rescues for partisan political advantage. While it is easy to deplore politically motivated attempts to influence Fed policy, we need to recognize the extent to which some measure of antagonism is an understandable consequence of the Fed's own credit policy initiatives.
The inevitable controversy surrounding central bank intervention in credit markets is one reason many observers have long advocated keeping central banks out of credit allocation.3 Central bank lending undermines the integrity of the fiscal appropriations process, and while U.S. fiscal policymaking may not inspire much admiration these days, it is subject to the checks and balances provided for by the Constitution. Contentious disputes about which credit market segments receive support, and which do not, can entangle the central bank in political conflicts that threaten the independence of monetary policymaking.
The independence that the modern central bank has to control the monetary policy interest rate emerged in stages following the end of World War II. The Treasury-Fed Accord of 1951 freed the Fed from the wartime obligation to depress the Treasury's borrowing costs. The collapse of the gold standard in the early 1970s and the attendant bouts of inflation led the Fed in 1979 to assert responsibility for low inflation as a long-term objective of monetary policy.4 The independent commitment of central banks to low inflation provides a nominal anchor to substitute for the anchor formerly provided by the gold standard.
The substantial measure of independence central banks were given was a key element in their relative success at sustaining low inflation over the last few decades. In fact, many other countries have adopted frameworks that hold their central banks accountable for a price stability goal, while allowing them to set interest rate policy independently in pursuit of their goals. This instrument independence has been critical to insulating monetary policymaking from election-related political pressures that can detract from longer-term objectives.
The cornerstone of central bank independence is the ability to control the amount of monetary liabilities it supplies to the public. But as a by-product, many central banks retain the ability to independently control the composition of their assets as well. For most modern central banks, standard policy in normal times is to restrict asset holdings to their own country's government debt. Some hold gold as well, a vestige of the gold standard. In addition, many make short-term loans to banks, either to meet temporary liquidity needs or as part of clearing and settlement operations; both are vestiges of the origin of central banks as nationalized clearinghouses.
The ability of a central bank to intervene in credit markets using the asset side of its balance sheet creates an inevitable tension. The desire of the executive and legislative branches to provide governmental assistance to particular credit market participants can rise dramatically in times of financial market stress. At such times, the power of a central bank to do fiscal policy essentially outside the safeguards of the constitutional process for appropriations makes it an inviting target for other government officials. Central bank lending is often the path of least resistance in a financial crisis. The resulting political entanglements, as we have seen, create risks for the independence of monetary policy.
This tension is a classic time consistency problem. Central bank rescues serve the short-term goal of protecting investors from the pain of unanticipated credit market losses, but dilute market discipline and distort future risk-taking incentives. Over time, small "one-off" interventions set precedents that encourage greater risk taking and increase the odds of future distress. Policymakers then feel boxed in and obligated to intervene in ever larger ways, perpetuating a vicious cycle of government safety net expansion.
The conundrum facing central banks, then, is that the balance sheet independence that proved crucial in the fight to tame inflation is itself a handicap in the pursuit of financial market stability. The latitude the typical central bank has to intervene in credit markets weakens its ability to discourage expectations of future rescues and thereby enhance market discipline.
Solving this conundrum and containing the impulse to intervene requires one of two approaches. A central bank could seek to build and maintain a reputation for not intervening, in much the way the Fed and other central banks established credibility for a commitment to low inflation in the 1980s. Alternatively, explicit legislative measures could constrain central bank lending. The Dodd Frank Act took steps in this direction by banning loans to individual nonbank entities. But Reserve banks retain the power to lend to individual depository institutions and to intervene in particular credit market segments in "unusual and exigent circumstances" through credit programs with "broad-based eligibility."5 In addition, the Fed can channel credit by purchasing the obligations of government-sponsored enterprises, such as Fannie Mae and Freddie Mac.
Constraining central bank lending powers would appear to conflict with the popular perception that serving as a "lender of last resort" is intrinsic to central banking. But even here, I think our historical doctrines and practices should not escape reconsideration. The notion of the central bank as a lender of last resort derives from an era of commodity money standard, when central bank lending in a crisis was the way to expand currency supply to meet a sudden increase in demand. Indeed, the preamble to the Federal Reserve Act says its purpose is "to furnish an elastic currency," not to furnish elastic credit. The Fed could easily manage the supply of monetary assets through purchases and sales of U.S. Treasury securities only.6 While it might sound extreme, I believe that a regime in which the Federal Reserve is restricted to hold only U.S. Treasury securities purchased on the open market is worthy of consideration.7
It might seem easy to criticize such a regime by reference to what it would have prevented the Fed from doing in the recent crisis. But that's the wrong frame of reference, I believe — it's an ex post, rather than an ex ante, perspective. Such a regime, if credible, would over time force changes in market practices that would alter the likelihood and magnitude of crises and the behavior of private market arrangements during a crisis. It would strengthen market discipline and incent institutions to operate with more capital and less short-term debt funding — changes we are now trying to achieve through regulatory means. The relative costs and benefits of such a regime may be difficult to map out conclusively. But I believe this tradeoff is well worth studying.
Ten years ago, my former colleagues Al Broaddus and Marvin Goodfriend argued that the design of central bank asset policy is "part of the unfinished business of building a modern, independent Federal Reserve."8 The 1951 Treasury-Fed Accord provided for independent Federal Reserve control of its liabilities, a necessary ingredient in monetary policy independence. But the accompanying power to use the Fed's asset portfolio to intervene in credit markets threatens to diminish that independence. Sorting out the conundrum of central bank asset policy should be high on the agenda for all those interested in improving the practice of central banking.
Jeffrey M. Lacker, "The Regulatory Response to the Crisis: An Early Assessment." Speech at The Institute for International Economic Policy and the International Monetary Fund Institute, Washington, D.C., May 26, 2010.
I am grateful to John Weinberg for assistance on these remarks.
Marvin Goodfriend and Robert G. King, "Financial Deregulation, Monetary Policy, and Central Banking," Federal Reserve Bank of Richmond Economic Review, May/June 1988, vol. 74, no. 3, pp. 3-12; Marvin Goodfriend, "Why We Need an 'Accord' for Federal Reserve Credit Policy: A Note," Federal Reserve Bank of Richmond Economic Quarterly, Winter 2001, vol. 87, no. 1, pp. 23-32; Robert L. Hetzel, "The Case for a Monetary Rule in a Constitutional Democracy," Federal Reserve Bank of Richmond Economic Quarterly, Spring 1997, vol. 83, no. 2, pp. 45-65; Marvin Goodfriend and Jeffrey M. Lacker, "Limited Commitment and Central Bank Lending," Federal Reserve Bank of Richmond Economic Quarterly, Fall 1999, vol. 85, no. 4, pp. 1-27; J. Alfred Broaddus, Jr., and Marvin Goodfriend, "What Assets Should the Federal Reserve Buy?" Federal Reserve Bank of Richmond Economic Quarterly, Winter 2001, vol. 87, no. 1, pp. 7-22.
Broaddus and Goodfriend, p. 8.
Such programs now require the approval of the Secretary of the Treasury.
The market supply of such securities is likely to be quite ample for some time to come. But even if the supply should shrink, as it did a decade ago, the Treasury could arrange to issue in sufficient quantities to allow the Fed to conduct monetary policy on a Treasuries-only basis. See Broaddus and Goodfriend.
See Goodfriend and King; Anna J. Schwartz, "The Misuse of the Fed's Discount Window," Federal Reserve Bank of St. Louis Economic Review, September/October 1992, vol. 74, no. 5, pp. 58-69; Goodfriend; and Broaddus and Goodfriend.
Broaddus and Goodfriend, p. 6.