João A.C. Santos and Stavros Peristiani
Though not literally a window any longer, the “discount window” refers to the facilities that central banks, acting as lender of last resort, use to provide liquidity to commercial banks. While the need for a discount window and lender of last resort has been debated, the basic rationale for their existence is that circumstances can arise, such as bank runs and panics, when even fundamentally sound banks cannot raise liquidity on short notice. Massive discount window borrowing in the immediate aftermath of the September 11 terrorist attack on the United States clearly illustrates the importance of a discount window even in a modern economy. In this post, we discuss the classical rationale for the discount window, some debate surrounding it, and the challenges that the “stigma” associated with borrowing at the discount window poses for the effectiveness of the discount window.
The Classical View
The classical view of the lender-of-last-resort (LLR) role begins with the work of Henry Thornton and Walter Bagehot in the late eighteenth and nineteenth centuries. Both authors stressed that the LLR was necessary not strictly to support individual banks, but to prevent a rapid fall in the money stock and to support the financial system as a whole during tumultuous times. To meet these broader objectives, Bagehot offered his famous dictum that the LLR should lend freely but at a high rate (relative to the pre-crisis period) to any borrower with good collateral, where good collateral included loans to private firms and bonds. The high rate was advisable, Bagehot reasoned, in part to limit demand from banks as well as to mitigate moral hazard—that is, to discourage banks from taking excessive risk in the expectation that the central bank would come to their assistance with emergency credit. When Congress created the Federal Reserve in 1913, it pointed out the need for “Federal reserve banks, to furnish an elastic currency, to afford means of rediscounting commercial paper, to establish a more effective supervision of banking in the United States, and for other purposes.” Two of these functions—furnishing an elastic currency and providing a means of rediscounting commercial paper—were important in leading the Federal Reserve banks to establish the discount window.
While the classical view of the LLR has shaped central banks’ policy responses during financial crises over the last century, it has also been debated. For instance, Bagehot’s other famous dictum, that central banks should lend only to illiquid but solvent banks, has been challenged. Some observers contend that central bankers are no better equipped to distinguish illiquid but solvent banks than are private investors. Others counter that central banks are better equipped because they may be privy to bank supervisory information to which private investors are not.
Goodfriend and King provide a spirited challenge to the view that central banks need a discount window. With the deepening of interbank markets since Bagehot’s day, perhaps emergency liquidity need only be extended to the market as a whole, through open market operations, rather than to individual banks. If banks themselves can distinguish solvent banks from insolvent banks, providing liquidity through open market operations (rather than to individual banks) would ensure the efficient allocation of liquidity from banks with surpluses to those with deficits while largely avoiding the moral hazard that might arise when the central bank lends to individual banks.
Notwithstanding the deepening of interbank markets, researchers have outlined a number of circumstances in which discount window lending is necessary to reduce the illiquidity of banks and the possibilities of panics occurring as a result of that illiquidity. For example, Flannery notes that discount window lending may be called for if and when—such as during a panic—a bank’s fundamental solvency may be hard for other banks in the interbank market to observe. Said differently, banks may be opaque even to other banks, particularly during a crisis. Diamond and Rajan point out that some banks may want to take advantage of “fire sales” during distress episodes, prompting them to hoard the liquidity provided through open market operations—an outcome that in turn leads to illiquid interbank markets. Gale and Yorulmazer observe that even apart from such opportunistic behavior, banks may hoard liquidity if they are uncertain about their ability to raise liquidity down the road.
If a bank facing a severe shortage of funds were not able to access either the interbank market or the discount window, it would be forced to rapidly liquidate loans and other assets. The premature liquidation of bank assets that might result is costly for the “real” economy, as it severs valuable relationships between banks and borrowers that may have been cultivated over years and destroys the information embedded in those relationships. Such relationships are not easily replaced, so the failure of a bank can cause financial distress to its borrowers, particularly smaller firms without access to capital markets. It also imposes a cost to the economy to replace the information that was lost when the relationship was abruptly severed. As importantly, the failure of one bank, particularly a large one, can trigger a banking system panic or lead to other bank failures by virtue of interbank exposures or other connections—a spiral that could culminate in a systemic failure with enormous costs to the economy. Discount window lending during recent financial crises may have helped avoid those costs. This aspect of the discount window recalls the Banking Act of 1935, which amended the Federal Reserve Act “to provide for the sound, effective, and uninterrupted operation of the banking system.”
While the discount window is an important tool for central banks dealing with liquidity problems that may threaten financial stability, its effectiveness depends critically on the willingness of banks to borrow from central banks. Banks are often reluctant to borrow from central banks not only because this source of liquidity tends to be expensive but also because of the “stigma” that is associated with discount window borrowing. As described in Courtois and Ennis, “stigma refers to a fear on the part of banks that a negative signal could be conveyed to regulators, other banks, or investors about a bank’s health if that bank is discovered to have borrowed from the Fed” (p. 1).
If a bank worries that borrowing from the discount window will lead other banks to doubt its fundamental solvency, it may avoid the discount window even if the discount window provides the cheapest funds available. Instead, the bank may liquidate marketable assets or try to borrow in the interbank market at onerous terms, further straining these markets and making it even more difficult for other banks to obtain funding or sell assets. Thus, central banks typically disclose only a limited amount of information about discount window activity to avoid branding healthy (but illiquid) banks as weak. The Federal Reserve, for example, has historically published the total amount of borrowing from the discount window on a weekly basis, but not information on individual loans. By allowing banks to borrow confidentially, this policy aims to make healthy institutions more willing to use the discount window during periods of market stress. It should be emphasized that confidentiality is not meant to protect the identities of individual banks per se, but rather to make the discount window more effective in dealing with market disturbances.
Central banks have long recognized the challenges that stigma creates for the effective operation of the discount window during crisis. Donald Kohn, former Vice Chairman of the Fed, has discussed the stigma problem in past speeches. “The problem of discount window stigma is real and serious. The intense caution that banks displayed in managing their liquidity beginning in early August 2007 was partly a result of their e
xtreme reluctance to rely on standard discount mechanisms,” Kohn noted in a 2010 speech. In fact, the need to mitigate stigma influenced the design of some of the lending facilities, such as the Term Auction Facility, created by the Fed during the financial crises.
In sum, the discount window is a vital tool to maintain the uninterrupted functioning of the banking system, but its effectiveness may be limited by the stigma associated with using it. This explains why policies that aim at dealing with the stigma of discount window borrowing are so important. Admittedly, the existence of the discount window may create some moral hazard, but of course, the Federal Reserve limits moral hazard by restricting discount window access to depository institutions that are closely regulated and supervised by federal banking authorities.
Below we provide some annotated references to papers on important aspects of the discount window. The Fed webpage on the discount window also provides much useful information.
I. Use of the Discount Window during Financial Distress
The discount window played a crucial role during several episodes of banking distress and financial disruptions (for example, the default of the Penn Central Corporation, the collapse of Continental Illinois, the massive problems in Treasury clearing settlements caused by the Bank of New York’s computer failure, the stock market collapse in October 1987, and September 11, 2001). In each of these episodes, the Federal Reserve supplied the needed liquidity, preventing a sharp deterioration in financial market conditions and insulating the economy from these shocks.
II. Early History of the Federal Reserve Discount Window
The discount window plays a key role in supporting Federal Reserve open market operations and the implementation of monetary policy. Since the establishment of the Federal Reserve in 1914, the role of the discount window has evolved as open market operations gradually replaced the discount window as the primary means of implementing monetary policy. The following articles provide good historical information on the discount window.
“Fundamental Reappraisal of the Discount Window,” prepared for the Steering Committee for the Fundamental Reappraisal of the Discount Window Mechanism appointed by the Board of Governors of the Federal Reserve System
III. Non-Price Mechanisms Governing Discount Window Borrowing
Prior to 2002, the discount rate was generally set below the overnight federal funds rate target. The positive spread between the competing money market rate and the discount rate engendered a profit incentive to borrow. To discourage profiteering, the Federal Reserve used non-price mechanisms, such as loss of potential future access, and other administrative restrictions rather than the discount rate to limit borrowing.
IV. The Reluctance to Borrow at the Discount Window.
Starting in 1980, the borrowing behavior of banks was affected by deteriorating financial conditions in the banking sector. Nevertheless, banks were increasingly reluctant to borrow from the discount window, fearing that market participants might interpret their borrowing as a signal of serious funding difficulties. This reluctance has become known as discount window stigma.
V. Recent Changes to the Discount Window
After the savings and loan crisis, Congress enacted the FDIC Improvement Act in 1991 (FDICIA) to recapitalize the Bank Insurance Fund. The legislation also mandated a prompt corrective action approach for the closure of insolvent banks. These new rules made discount window access more restrictive for undercapitalized banks.
In 2002, the Board of Governors established two new discount window programs (primary credit and secondary credit) to replace the existing facilities. Primary credit is advanced to banks in sound financial condition with short-term needs. More important, the new rules adopted a penalty borrowing rate that set the interest rate on primary credit above the federal funds rate.
The views expressed in this post are those of the author(s) and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Any errors or omissions are the responsibility of the author(s).