Eliminating Discount Window Stigma: What Can We Learn from Abroad?

This article is the second in a three-part series on stigma and discount window borrowing and picks up from an earlier article on discount window design to see how the experiences of other central banks can inform work to redesign the discount window to reduce stigma. As explained in that article, banks’ reluctance to use the discount window is problematic for financial stability as it constrains the Fed’s ability to use its liquidity provision tools to stem runs and mitigate contagion in times of stress. The stigma associated with discount window borrowing in the US is well documented and is a multifaceted phenomenon, as described in another previous YPFS article.

This YPFS article examines the liquidity provision tools of four other central banks—the Bank of Canada (BoC), the Bank of England (BoE), the European Central Bank (ECB), and the Swiss National Bank (SNB)—to explore lessons from their liquidity tools that could be applied to make the Fed’s lender of last resort tools more effective and less stigmatized. These central banks were chosen both for the transparency of their operating frameworks and because little or no stigma is associated with their on-demand liquidity provision tools.

How do we know that these central banks’ standing liquidity facilities are not stigmatized? In the reserves scarcity frameworks many central banks used to implement monetary policy prior to the Global Financial Crisis (GFC), standing facilities were designed to get reserves to banks that weren’t able to obtain them in the market on a given day. They were priced to put a ceiling on money market rates, as a means to limit rate volatility and support interest rate control. Stigma was evident when banks preferred to pay a higher rate to borrow in the market to avoid using the central bank’s standing facility.[1] But at present, just like the Fed, the BoC, BoE, ECB and SNB are all operating with abundant levels of reserves, with relatively little demand for central bank liquidity except in periods of financial market stress, making usage-based assessments of stigma difficult. To determine whether these central banks’ facilities experience the type of stigma associated with the discount window, I reviewed the few academic pieces that have been published on the topic, and spoke with staff at each central bank about their assessment of stigma based on their ongoing market monitoring.[2] In all four cases, staff confirmed that they saw no evidence of stigma that would deter banks from using the central bank’s standing facilities when needed, based both on their ongoing conversations with banks and banks' willingness to use standing facilities in recent episodes of funding market stress.

How are other central banks’ liquidity provision tools designed?

Tables 1 through 5 summarize publicly available information on the current BoC, BoE, ECB and SNB operating frameworks for liquidity provision to the banking sector. The focus here is provision of local currency liquidity, on an overnight basis, and excludes cross-currency tools such as central bank swaps. Table 1 shows the tools each central bank has to provide liquidity to banks, categorized as monetary policy tools, liquidity backstop tools, recovery funding and, where the central bank plays a role in it, resolution funding. Table 2 summarizes how these tools are priced. Table 3 highlights several features of these frameworks for liquidity provision that contrast with the Fed’s framework. Table 4 illustrates how these central banks communicate about their liquidity provision tools. And Table 5 summarizes how supervisors in each jurisdiction treat capacity to borrow at the central bank in their oversight of banks’ liquidity risk management. 

* per Article 55 of EU Guideline 2015/510 of the ECB

** CAMELS are composite supervisory ratings of bank condition; 1 is the highest rating, 5 the lowest.

*** Subject to limits on lending to undercapitalized institutions under the FDIC Improvement Act of 1991.

* STLF rate can never be less than the Bank rate, equal to OIR + 25, according to Bank of England website

** The ECB has announced that the MLF rate will be reduced to DFR + 40 effective September 18, 2024

*** LSFF rate can never be less than 0%, according to Swiss National Bank website

Observations from cross-jurisdictional comparisons

Operating framework design affects the level of stigma associated with the central bank’s liquidity provision tools. As shown in Table 1, the BoC, BoE, ECB and SNB all draw a much brighter line between on-demand lending to healthy banks and discretionary lending to troubled ones. Clear separation helps to avoid stigmatizing use of the on-demand facility. By contrast, as argued in a previous article, the Fed’s administration of programs for lending to both healthy and troubled banks under a single “discount window” appears to have stigmatized primary credit by association with secondary credit.

Differences in how monetary policy is implemented also matter for stigma. The US monetary policy framework does not rely on liquidity provision to banks in periods of policy accommodation—unlike other jurisdictions—for two reasons. First, in recent years, collateralized lending has played a smaller role in US liquidity-providing operations (relative to asset purchases) than in other jurisdictions. The BoC, BoE, ECB and SNB relied heavily on term repos with banks to inject liquidity during the pandemic, compared with the Fed which relied more heavily on large scale asset purchases from primary dealers. Second, while banks are counterparties for monetary policy operations of all four peer central banks analyzed above, very few banks meet the eligibility requirements to participate in the Fed’s monetary policy operations. As a result, bank borrowing from the central bank has been more commonplace, and thereby less stigmatized, in recent years in Canada, the eurozone, Switzerland, and the UK. Banks in the US thus appear to be less incentivized to pre-position collateral at the central bank than those operating in jurisdictions where pre-positioned collateral can be utilized in monetary policy operations.

The central bank’s communications about its liquidity tools and framework materially affect the level of stigma. In Canada, the eurozone, Switzerland, and the UK, the central bank’s public stance is that short-term borrowing from the standing facility is a perfectly acceptable source of liquidity for solvent banks and is available on-demand. This stance is mirrored in supervisory messaging to banks. Notably, neither of these things have consistently been true in the US since the current incarnation of the discount window was launched in 2003. That redesign was initially accompanied by clear and repeated communication to banks and the general public that primary credit borrowing was desirable—both as a mechanism to support interest rate control for monetary policy and as a tool to ensure distribution of reserves to where they were needed—and that it would be granted on a “no questions asked” basis. However, this messaging faded amid criticism of the Fed’s lending to banks (and nonbanks) during and after the GFC. While Fed policymakers have recently been making a concerted effort to urge that all banks be ready to use primary credit, and that they should use it when needed,[3] many banks continue to report that their supervisors and/or senior management continue to discourage use. Indeed, the May 2023 version of the OCC supervisory handbook on liquidity still contains language casting doubt on its availability when needed:

The discount window is available to relieve liquidity strains for individual banks as well as the banking system, but the Federal Reserve Banks are not required to lend through the discount window.

As discussed in a previous article, this language reflects the so-called “constructive ambiguity” about discount window lending that was built into Section 10B of the Federal Reserve Act with the passage of the FDIC Improvement Act of 1991. This language appears to reflect concerns about lending to failing institutions—but does not distinguish between sound and weak banks and thus applies uncertainty to all discount window lending. Such uncertainty is not present in other central banks’ communication about their on-demand standing facilities and is not consistent with the idea of routine access to liquidity—contributing to stigmatization of primary credit in the US.

Communication about eligibility for the routine, on-demand facility can affect stigma. The Fed describes primary credit as a tool for “generally sound” banks, and secondary credit as a tool for “banks not eligible for primary credit”: implying that “unsound” banks can borrow from the discount window. Primary and secondary credit eligibility are determined by the composite CAMELS rating of the bank, leaving little room for judgment about lending to a bank whose condition has deteriorated materially since its last supervisory examination and rating update. This can be viewed as a more comprehensive, and in theory more effective, approach to assessing a bank’s safety and soundness than simply focusing on capital adequacy. But reliance on a numerical rating is only as sound as the timeliness of that rating. CAMELS ratings are not updated frequently; Silicon Valley Bank and Signature Bank are only the latest examples of banks that had a composite rating of 3 or better until shortly before failure. Indeed, the failure of banks that were eligible for primary credit until just before their demise may itself have contributed to the stigma associated with primary credit.

By contrast, the BoC, BoE, ECB and SNB all simply state that the on-demand standing facility is available only to solvent banks, while not defining solvency precisely. This approach retains room for judgment in lending decisions that can help to promote financial stability, while also signaling to the public that use of the on-demand standing facility is not to be interpreted as a sign of distress at the borrowing bank, thereby mitigating stigma.

Including central bank borrowing capacity in supervisory measures of contingent liquidity resources helps to reduce stigma. While the Fed worked closely with supervisors in the years immediately following the 2003 redesign of the discount window to align on the “no questions asked” messaging on primary credit usage, the new liquidity and resolution planning requirements put in place in the US after the GFC generally did not count primary credit borrowing capacity toward a bank’s contingent liquidity resources. Table 5 summarizes how central bank borrowing capacity is currently treated in the supervisory frameworks of Canada, the UK, the eurozone, and Switzerland. A comparison to the US shows that some other jurisdictions have gone a bit further to integrate central bank liquidity provision into the regulatory view of banks’ options for contingent liquidity management. For example, some jurisdictions permit a bank to count its capacity to borrow at the routine standing facility (as measured by pre-positioned collateral value) in its internal stress tests and/or resolution funding plans. And some require all eligible banks to test usage of the standing facility regularly, as US banking regulators are now considering.

Some features of the US institutional framework for liquidity provision pose unique challenges for stigma reduction that other central banks do not face. For example, the Fed is required by law to publish the details of all discount window lending transactions, including borrower and amount, at a two-year lag; this disclosure is widely viewed as a factor driving discount window stigma. No other central bank makes individual borrower information available to the public. Moreover, the format of the Fed’s weekly publication of balance sheet data causes certain borrowers to be identified by the market within a week of borrowing.

A second feature is the Federal Home Loan Bank (FHLB) System. FHLBs are government-sponsored enterprises that were established in 1932 to provide liquidity to banks to support housing finance and community investment. The FHLB system has grown over time to become a large-scale provider of funding to banks beyond the scope of housing finance, often on terms more favorable than those at the Fed’s discount window. The FHLBs are viewed by many banks as a preferable alternative to the Fed for obtaining funding in size, due to their ability to lend at longer terms than the discount window and their relative operational efficiency and pricing.

A penalty rate does not necessarily create stigma, if the central bank’s standing liquidity facility is designed and communicated to mitigate it. Table 2 shows variability in practice across central banks with respect to pricing of routine liquidity facilities. Some central banks tier interest rate pricing based on collateral type; some add a premium in pricing term lending while others do not. Notably, the BoC, BoE, ECB, and SNB all price their liquidity facilities at varying spreads above the policy rate, yet observe little or no stigma associated with their usage. In contrast, the Fed’s highly stigmatized primary credit program is currently priced right at the top of the policy target range. Factors other than price clearly matter for stigma.

Policy recommendations

The BoC, BoE, ECB, and SNB have all established effective standing liquidity tools that their banks are willing to use when needed. Since the GFC, the Bank of Canada and the Bank of England have both revised their frameworks for liquidity provision to overcome the stigma associated with earlier tools. They did so through clear, ongoing public communication about the framework (as shown in the excerpts in Table 4) and by taking steps to align supervisory expectations with the policy message about the legitimacy of use of standing facilities (as summarized in Table 5).

As shown in Table 4, the ECB and the SNB both explicitly describe the on-demand standing liquidity facilities for healthy banks as being part of their overall monetary policy implementation frameworks. Their experiences suggest several actions the Fed could take, in partnership with the other bank regulators, to establish an on-demand liquidity facility that is not stigmatized.

1. Separate the administration and governance of routine, on-demand lending to sound banks from lending to weaker institutions to support restructuring or resolution.

The experiences of other central banks suggest that drawing a brighter line between funding for healthy banks and funding for weak banks with respect to public communication, program administration, and governance could help to avoid the stigmatization of tools for lending to healthy banks by proximity to tools for lending to weaker banks.. Redesigning and rebranding the Fed’s tools for liquidity provision in a manner that clearly distinguishes recovery and resolution funding for troubled banks from a new on-demand standing liquidity facility for sound institutions is essential to ensure the on-demand facility is not stigmatized.

2. Communicate clearly and resolutely to the public that use of the new on-demand liquidity facility is appropriate for healthy banks —through both policy and supervisory channels.

As noted above, the Fed had some success doing this at its launch of the primary and secondary credit programs in 2003. The Fed could leverage its dual role as liquidity provider and bank regulator to pursue such alignment again. Aligning supervisory and policy messaging to banks that the standing facility is “open for business,” as the Bank of England describes their liquidity provision tools, is crucial to overcome the learned views of bankers (and those who invest in or write about them) that borrowing from the central bank is to be avoided at all costs. Doing so will make the Fed’s liquidity provision toolset much more effective at safeguarding financial stability in stress.

3. Recalibrate supervisory requirements to support a new routine, on-demand lending tool without stigma.

Aligning regulatory and supervisory requirements for banks’ liquidity risk management with the idea that the Fed’s standing liquidity facility for sound banks is a legitimate source of contingent liquidity would go a long way to reducing stigma. Some other jurisdictions allow banks to count their capacity to borrow at the central bank’s on-demand facility in their internal stress tests and/or resolution plans; the US should do this as well. And as noted above, the institutional framework for liquidity provision in the US has some features that amplify stigma that other central banks do not face. Reducing stigma might thus require the US to go further than others must to incentivize bank usage of central bank liquidity facilities when needed, to compensate for these challenges. For example, supervisors could count some fraction of a sound bank’s pre-positioned loan collateral in the numerator of its Liquidity Coverage Ratio, on the grounds that this collateral could be monetized on a same-day basis at the standing facility, or include it in the denominator of the ratio as an expected inflow from borrowing from the routine, on-demand facility.

4. Evaluate whether including banks more fully in the Fed’s counterparty framework for monetary policy operations could help to avoid stigmatization of the new on-demand standing liquidity facility for healthy banks.

As noted above, in other jurisdictions banks are counterparties to the central bank in its monetary policy operations and can obtain liquidity in repo operations in periods when the central bank is expanding its balance sheet. Integration of bank borrowing into the monetary policy framework can help to reduce the stigma associated with central bank borrowing. An evaluation of the extent to which including banks more fully in the Fed’s counterparty framework for monetary policy operations could help to mitigate stigma in periods of policy easing—for example, by creating a new tier of counterparties for repo operations, as has been done for reverse repo operations—would be worthwhile. While banks are eligible to use the standing repo facility (SRF) that was established in 2021 to support repo market functioning, they have generally not used it due to the above-market rate. In contrast, repo operations to implement monetary policy are conducted through an auction in which all primary dealers are required to participate, and which results in an auction-determined price.

5. Improve the timeliness of assessments of bank condition, and communicate publicly about eligibility for the on-demand facility accordingly.

Perhaps the greatest difficulty a central bank faces as a lender is determining whether a bank is eligible for routine, on-demand liquidity or not. The “solvency” standard that other central banks use to determine who can use their on-demand liquidity facility is broad enough that it provides some room for judgment amid uncertainty about whether the bank is truly a going concern or not.

In contrast, use of a very specific metric such as the composite CAMELS rating—and transparency to the public about this decision rule—carries two risks. The first risk is that reliance on a very specific metric may constrain judgment about, and flexibility to address, a bank’s situation in a stressed market environment. If supervisors downgrade the composite rating below a 3, the central bank must mechanically deny the bank access to primary credit, regardless of the reason for the downgrade or prospects for recovery. A bank that becomes undercapitalized should of course be removed from primary credit, per the FDICIA restrictions on lending to undercapitalized banks. But CAMELS downgrades can occur due to downgrades in component ratings other than capital adequacy, and there might be situations in which a mechanical denial of primary credit might not be the right action to take from a financial stability standpoint. The second risk is that given the time lag between supervisory ratings updates and the speed at which a bank’s condition can change materially, primary credit can be extended to a bank that is truly troubled, prolonging its life when it should in fact be closed.

As a new on-demand liquidity facility for healthy banks is implemented, eligibility for it should be communicated carefully, in a manner that reserves some decision space for policy makers to make the difficult determination in real time that a borrower is no longer healthy, as other central banks have done. It would also be advisable to supplement CAMELS ratings that are often stale with more timely judgments about current and future health when a bank comes under pressure, to support the central bank’s lending decisions. The work now being undertaken in the US to improve supervisory processes may yield improvements on this front.

Conclusion

Amid growing debate about the currency of the discount window infrastructure and suggestions by Fed officials that they are looking to support financial stability by modernizing and destigmatizing discount window lending operations, the time seems ripe for consideration of design improvements to address the important issue of stigma. The experiences of the Bank of Canada, Bank of England, European Central Bank, and Swiss National Bank suggest that it is possible to do so—with the right framework design; the right communication; and good alignment between policy, regulatory and supervisory messages. Fortunately, the design, communication, and supervisory changes that are needed to establish an on-demand lending tool that healthy banks can and will use lies largely within the Fed’s and the other banking regulators’ existing authorities. 

[1] See for example https://www.chicagofed.org/publications/working-papers/2004/2004-01 for empirical evidence of discount window stigma in the U.S., demonstrating banks’ willingness to pay up in the market rather than obtain a cheaper rate at the window.

[2] See https://libertystreeteconomics.newyorkfed.org/2018/04/is-stigma-attached-to-the-european-central-banks-marginal-lending-facility/ for an assessment of stigma in ECB standing facilities. See also https://libertystreeteconomics.newyorkfed.org/2016/09/is-there-discount-window-stigma-in-the-uk/ for an assessment of the impact of post-GFC reforms of the Bank of England’s standing facilities on stigma.

[3] See for example, Vice Chair Barr’s May 20 speech and President Logan’s May 10 reported remarks.