Bill Nelson on the Using the Discount Window for Liquidity Requirements and Its Implications for the Fed's Balance Sheet

Episode Summary

Executive Summary: Bill Nelson argues that post-SVB regulatory reform should make the Fed’s discount window a normal, pre-positioned source of liquidity that banks can count in supervision, reducing stigma and reliance on large reserve buffers. He also says this could help shrink the Fed’s balance sheet, but only if rules, Fedwire hours, and supervisory attitudes change together.

Main Topics: Reforming the discount window as usable liquidity (Priority: 5/5): Nelson supports proposals requiring banks to hold collateral or access to liquidity sufficient for near-term cash outflows, and to have that capacity count in liquidity supervision rather than being ignored. Stigma and the history of discount-window borrowing (Priority: 5/5): He traces the window’s stigma to the 1920s and especially the post-2008 era, arguing that supervisory and public treatment turned routine borrowing into a sign of weakness. Liquidity regulation and the role of collateral (Priority: 4/5): Nelson says the post-GFC shift toward holding only HQLA pushed banks toward treasury-like assets and away from lending, while pre-positioned loan collateral could preserve credit creation and improve resilience. Fed balance-sheet size and the reserve ratchet (Priority: 5/5): He argues that abundant reserve regimes create a ratchet effect: each QE episode raises banks’ and supervisors’ expectations for reserve buffers, making it harder to shrink the Fed’s balance sheet. Critique of the floor system and reserve-demand forecasts (Priority: 4/5): Nelson says the 2018 framework was sold as simpler, but in practice it required large reserves, ongoing fine-tuning, and later massive repo operations after the September 2019 stress. Fed operating losses and interest-rate risk (Priority: 4/5): He discusses 2023 losses, stressing they do not stop the Fed from operating, but they raise questions about whether large balance-sheet expansion was worth the risk and cost to Treasury remittances. Operational improvements: Fedwire and the standing repo facility (Priority: 3/5): He suggests 24/7 Fedwire and better integration of the standing repo facility into contingency planning would improve liquidity management and reduce stress-period frictions.

Key Arguments: Banks should be judged on their ability to meet short-term outflows using pre-positioned collateral at the discount window, not just by holding large quantities of HQLA. The discount window is a normal central-bank tool; treating borrowing as exceptional or shameful undermines financial stability. Using loan collateral at the window supports both safety and credit supply, unlike forcing banks to hold only treasuries or reserve balances. Liquidity rules and supervisors should allow banks to rely on the discount window and standing repo facility in contingency planning. The Fed’s oversized balance sheet and floor system have created a ratchet that makes balance-sheet reduction difficult and potentially unnecessary large. The 2018 claim that abundant reserves would simplify implementation proved overly optimistic after the 2019 repo episode. Fed losses matter politically and fiscally because they reduce remittances to Treasury, and large reserve-based funding makes losses more likely when rates rise. Fedwire’s limited hours are outdated and can worsen liquidity strains in modern markets. Any change to this framework should be done through formal rulemaking, not informal examiner pressure.

Data Points: SVB deposit runoff: About three-quarters left in 36 hours - Used to illustrate how rapidly deposits can flee under stress and why short-term liquidity preparedness matters. Discount-window collateral estimate: $1.6 trillion - Fed’s last published estimate of collateral already pledged or available for the window, though Nelson thinks it is likely higher now. Collateral composition at the window: About 80% loans - Nelson says most pledged collateral is loans rather than securities. Consumer-loan collateral share: About 40% - Part of the collateral mix at the discount window. Business-loan collateral share: About 20% - Part of the collateral mix at the discount window. Fed reserves before the crisis-era regime: About $10 billion - Reserve balances under the older scarcity/clearing framework. 2008 staff estimate for ample reserves: $35 billion - April 2008 estimate of the reserves needed under the old-style operating approach. 2016 staff estimate: $100 billion - March 2016 estimate as reserve demand began to rise. 2018 staff estimate: $600 billion to $800 billion - March and December 2018 estimates around the shift to the floor system. 2019 staff estimate: $1.3 trillion - September 2019 estimate after the framework shift. 2022 staff estimate: $2.3 trillion - May 2022 estimate showing the reserve ratchet continuing upward. Current reserve balances: About $3.5 trillion - Nelson’s estimate of reserve balances in the current system. Fed balance sheet peak: $8.9 trillion - Peak of the Fed’s balance sheet mentioned in the discussion. Fed balance sheet current size: $7.6 trillion - Approximate size after some QT reduction. Treasury holdings peak/current: From about $5.8 trillion to $4.7 trillion - Fed Treasury holdings have declined from their peak. Overnight reverse repo peak/current: From about $2.5 trillion to under $1 trillion - RRPs have fallen substantially during QT. 2023 Fed operating loss: $114 billion - Annual loss reported by the Fed. Potential foregone seigniorage/profit: About $233 billion - Nelson says the Fed effectively lost $114 billion plus the $119 billion it could have earned from simple bill investment. Hypothetical annual profit from currency invested in T-bills: $119 billion - Counterfactual used to show the opportunity cost of the Fed’s current funding structure. Weekly loss rate: About $2.25 billion to $2.5 billion - Approximate weekly pace of Fed losses mentioned in the interview. Discretionary spending comparison: About one-seventh - Nelson compares $233 billion to a seventh of federal discretionary spending. Low-reserve policy concern: $300 billion above unknown minimum - Previous Fed plan to shrink until reserves were about $300 billion above the unknown minimum necessary level. Chair Powell concern in 2018: $1.3 or $1.5 trillion - Powell said he’d have buyer’s remorse if the needed reserve level proved this high rather than $1 trillion.

Pivotal Quotes: "The discount window is available right immediately, and it's available up until the end of, you know, well after the end of the day." — Bill Nelson: Explaining why the window is uniquely suited for emergency cash-flow needs. "If you allow them to do what the Federal Reserve was designed to do back in its founding and allow those banks to make loans to businesses and households and pledge those loans to the discount window as collateral to establish lendable value as your response, then you're making the system safer and more liquid." — Bill Nelson: Core argument for using the discount window as a liquidity backstop without shrinking bank lending. "The reason why the standing repo facility, it was created in part because of concerns about repo markets and providing credit to primary dealers. But the other role, the role as a stigma-free discount window for banks, that's entirely about packaging." — Bill Nelson: Clarifying that SRF and discount window are conceptually similar for banks, but the branding and operational design differ.

Implications: If regulators normalize discount-window use and allow collateral-based liquidity to count in supervision, banks may need fewer reserves, the Fed could shrink more safely, and crisis response would improve. But this requires formal rulemaking, better hours, and a major shift in supervisory culture.