Recent experience, particularly with the Bank Term Funding Program, shows how to improve the functioning of the Fed’s discount window and strengthen financial stability.
By Jeff Huther and Alison Touhey
ABA Viewpoint
Over the years, there have been many proposals to improve the Federal Reserve’s discount window. In recent months, U.S. bank regulators (here and here) have highlighted their interest in addressing the window and banks’ use of this facility.
As policymakers and stakeholders revisit past discussions on this topic and reflect on insights from the now wound-down Bank Term Funding Facility, ABA is putting forth some initial recommended changes to the discount window. These include enhancing the window’s desirability by improving its operational ease, updating its collateral pricing and better integration of the window into supervisory oversight. A striking aspect of most of these recommendations is that their implementation would not require statutory or regulatory changes.
In this article we expand on ABA’s recent statement to highlight various related supervisory and policy questions that need to be explored to reposition the window as a usable liquidity backstop. These include regulatory expectations on when the discount window should be used, the optimal levels of usage and whether the window is only a short-term emergency lending facility or if it can also be used as a business-as-usual liquidity source to supplement and manage funding flows. Efforts to improve the discount window should not undermine or disadvantage other market-driven critical sources of liquidity, such as the Federal Home Loan Bank system, or lock up high-quality collateral at the Federal Reserve.
The Fed, together with the OCC and the FDIC, should use the current period of calm to review and improve the discount window. Statutory and regulatory changes should be evaluated within the broader context of initiatives related to liquidity supervision and deposit insurance, among others. Recent speeches by officials show that they are aware of the need for change, and ABA has submitted a letter to the banking agencies in response to statements by and on bank liquidity risk management and potential new liquidity regulations.
Improve operational ease
Established in 1913, the discount window was conceived to be the ultimate source of emergency liquidity for the banking industry. Banks could use this facility as a source of short-term, secured funding from the Federal Reserve. In theory, the window can smooth a variety of liquidity outflows due to seasonal factors, bank-specific financing needs, or systemic stress affecting conditions out of the bank’s control. In practice, its use is hindered by simple operational hurdles as well as concerns about usage being seen as a sign of desperation.
As a threshold matter, the current operating hours of 8:30 a.m. to 7 p.m. ET are ill-suited to meeting banks’ liquidity needs in today’s global, 24/7 operating environment. Fedwire’s hours of operation should be expanded to support the global dollar-based financial system. The Fed’s price structure for the Fedwire, which includes a small fixed fee and a per-transaction fee that is lower for larger numbers of transactions, penalizes smaller banks with fewer transactions. While the Fed has noted that small banks were more frequent discount window users from 2019 to 2022, their “small bank” category included banks with up to $3 billion in assets. The pricing structure should be re-evaluated so that banks of all sizes are incentivized to transact with the Fed. This process should include a consideration of consistency (of the Fed’s pricing structure) with the policy goals of the regulatory authorities.
The Fed should also streamline the setup process and align the collateral eligibility and management processes with market standards; doing this will also reduce operational bottlenecks. Further, as widely recognized, the Federal Reserve and the FHLBs should work together to ensure that collateral can be transferred smoothly under tight timeframes.
Update pricing of collateral
One of the defining characteristics of financial panics is that market participants, for any number of reasons, drive market prices below fundamental values. As financial conditions tighten, the value of most collateral falls, making it more difficult for borrowers to obtain the necessary amounts of liquidity. By relying on market prices, the discount window exacerbates liquidity stress by forcing banks to post increasing quantities of collateral to borrow a certain amount. To meet the central bank’s goal of calming stressed markets, the Fed should be lending against the underlying values of the collateral, rather than the panic-induced market prices.
A related issue is that the steep discounts (also referred to as haircuts or margins) on collateral that protect the Fed from unlikely credit losses are procyclical and discourage banks from using the window as a source of funding in all but the most extreme circumstances. One question for the Fed to consider is to how much credit risk the window should be exposed to, especially given that discount window borrowing tends to be overnight in nature. The existing collateral discounts appear to be set to ensure the Fed faces no credit losses in all but the most extreme situations — a setting that may be inconsistent with the Fed’s broader financial stability goals and in stark contrast to its willingness to face interest rate losses.
The Fed’s modeling of collateral value may also be discouraging discount window participation. The current setting of collateral discounts is opaque and can vary over time without explanation. Transparency around the collateral valuation models would be a small but helpful improvement; even better would be a clear rationale for the assumptions applied in the models. Even if that rationale does not lead to valuation stability, added information regarding the drivers of the change would avoid unexpected movements in prices.
While the bank stress of March 2023 is not representative of many historical bank failures, it is worth considering that Silicon Valley Bank’s weaknesses were magnified by its inability to borrow against the par, rather than market, value of its Treasury holdings — a point that has been reinforced by the Fed’s creation of the BTFP which, in contrast to the discount window, provided loans against the par value of Treasury securities. The design of the BTFP program clearly reflected a consideration that credit risk concerns should be overridden during stressed market conditions. We encourage the Fed to rethink its collateral policies to make them consistent with the policy goals of the discount window.
Lessons from the BTFP
The creation of the BTFP (and its predecessors in prior stress events) was implicit recognition that the discount window could not provide sufficient liquidity quickly, efficiently and at a reasonable cost. The BTFP, measured by participation, was a success — which suggests that application of BTFP-type terms could increase discount window acceptance. At times, the BTFP rate was better than market alternatives, making it difficult to identify what other BTFP features would improve DW acceptance. Experimentation with other BTFP features has little downside since DW usage is generally close to zero. BTFP parameters that could be applied to discount window loans include offering loans based on the par value of high-quality collateral, eliminating haircuts, experimenting with maturity terms and considering alternative interest rates.
-
- Par value. One of the factors that made the BTFP so successful is that it did not cap the amount of Fed lending at the discounted market value of eligible collateral. Instead, the BTFP accepted bank holdings of high-quality collateral at par value, so as not to penalize banks for holding high-quality assets for prudential liquidity management. The routine use of par value for high-quality collateral, such as Treasuries, should be studied with full awareness that it would actually decrease borrowing capacity when interest rates have fallen (because market value is generally greater than par value in these cases). In any case, the Fed’s provision of backstop liquidity policy based on fluctuating prices is worth reviewing.
-
- Haircuts. In addition to limiting loans to the market value of a bank’s collateral, the window also discounts the market value of the collateral when determining the bank’s borrowing capacity. The BTFP parameters did not include haircuts that may have contributed to the initial success of the program. The rationale for discount window haircuts should be re-examined as part of modernization and any re-evaluation of exposure to credit risk.
-
- Maturity terms. Discount window loans, by statute, must have maturities of less than four months. In practice, almost all loans are repaid the next business day. The success of ad hoc alternatives such as the Term Auction Facility and BTFP may have been due, in part, to longer terms. It is possible that the stigma associated with discount window borrowing could be reduced if the Fed offered standing term loans. A systematic approach to window loans of, say one or three months, may reduce the perception that activity at the window is only conducted in urgent, crisis-type situations.
-
- Interest rates. An extensive review of discount window operations in 2002 (summarized here) led to a price-based approach to window use. Over the 40 years prior to the 2003 revisions, the rate had generally been set at or below the Fed’s policy target rate. Arbitrage-driven usage was discouraged administratively. Since 2003, usage has been discouraged by penalty pricing, initially 100 basis points over market rates, reduced to 50 basis points over market rates in 2008 and around 15 basis points since the pandemic. This pricing is consistent with the conventional view that central banks should lend against good collateral at high rates during periods of stress. Recent proposals, however, have suggested a more prominent role for the discount window during normal times. Those proposals should recognize that more active use of the window requires an interest rate that is not uneconomical for users. If the goal is for banks to use the discount window during normal times, then why penalize them through a higher rate?
Give explicit supervisory credit for discount window use
Banks’ historical reluctance to borrow from the window can be attributed to a combination of operational hurdles, supervisory discouragement and concerns that usage signals desperation (embedded in the label “lender of last resort”). Some combination of these factors has led to the view that the window is stigmatized, in that banks do not use the window even when it is notionally cheaper than alternative funding sources. The stigma problem was exacerbated by the Dodd-Frank Act’s requirement that the Fed disclose discount window borrowers with a two-year lag.
While the vague nature of the stigma makes it impossible to create a menu of specific steps needed to destigmatize the discount window, supervisors can take simple steps that would unquestionably help in the near-term. The recent supervisory emphasis on discount window readiness will have a positive effect on stigma reduction. Further, for supervisory purposes, banks that have regularly tested open lines, for example, should garner a higher “L” rating in CAMELS and LFI ratings.
Borrowing from the window is considered by many stakeholders — including bank executives, investors and supervisors — to be a sign of weakness. The acceptability of window use has to be absorbed by all stakeholders, not just supervisors. Bank executives, for example, are unlikely to make greater use of the window if supervisors accept usage as normal course of business but bank investors do not (or vice versa). We appreciate Barr’s comments in his recent speech that “using the discount window is not an action to be viewed negatively.” We urge the Fed to continue educating banks, supervisors, investors and the public; though it is debatable if market participants’ view on the window will change simply based on such comments in public speeches.
There is ongoing debate about the appropriate level of DW usage and how it interacts with other funding providers, such as the Federal Home Loan Banks. The debate should rule out banks always turning to the Fed for funding and banks never turning to the Fed. Always is undesirable because our financial system is more robust when banks can borrow and lend in deep private markets. Never is undesirable because economic shocks are inevitable, and centuries of financial and economic history show that without a deep liquidity backstop, banks facing those shocks can magnify adverse economic outcomes. In between always and never lie judgment calls, changing market conditions and the effects of regulatory changes on bank balance sheets that may be altering supervisory views over time. The creation of the BTFP in March 2023 suggests that officials recognized a need for greater Fed lending but concluded that the discount window, as currently structured, would not satisfy that need.
Conclusion
The Federal Reserve was established in 1913 to provide liquidity to the banking system, particularly in times of economic or financial stress. To fill the gap between the window and the needs of modern banks, the Fed has resorted to ad hoc institutional structures to address instability in short-term funding and other debt markets during stressed market conditions. Barr’s recent speech suggests that the Fed has not given up on returning the window to its original purpose. To do so, however, fixes like the ones described above, along with a rethinking of the policy goals, are needed to bring the discount window into the 21st century.
The Fed has the authority to make operational changes within its existing regulations and statutes. It should use that authority where appropriate to experiment with parameters that meet its stated goals. Modernizing discount window operations should take priority over writing new rules with questionable policy goals and market outcomes.
Jeff Huther is VP and senior economist at ABA. Alison Touhey is SVP for bank funding policy at ABA.
ABA Viewpoint is the source for analysis, commentary and perspective from the American Bankers Association on the policy issues shaping banking today and into the future. Click here to view all posts in this series.
