George Selgin on Repo Market Stress, Fed Balance Sheet Volatility, and a Standing Repo Facility

Episode Summary

Executive Summary: The episode examines why the Fed’s September repo-market spike happened and why similar stress could recur at year-end. George Selgin argues that falling reserves were driven not only by QT, but also by volatile Fed liabilities—the Treasury General Account and the foreign repo pool—which drained reserves from banks. He proposes a standing repo facility plus changes to reduce and stabilize those liabilities.

Main Topics: What happened in the September repo spike (Priority: 5/5): The discussion begins with the September 2019 surge in repo rates to nearly 10%, showing that reserves were too scarce for the funding needs of the repo market and forcing the Fed to inject liquidity through repos and asset purchases. Fed balance sheet liabilities that affect reserves (Priority: 5/5): Selgin explains the liability side of the Fed’s balance sheet, emphasizing that beyond currency and bank reserves, the Treasury General Account and the foreign repo pool can absorb reserves and reduce what banks have available. How QT and non-reserve liabilities drained reserves (Priority: 5/5): Quantitative tightening reduced reserves directly, while TGA growth and foreign official deposits at the Fed indirectly shifted reserves out of the banking system, making reserve scarcity much worse. Why the TGA became volatile after 2008 (Priority: 4/5): The Treasury shifted from keeping minimal balances at the Fed to using the TGA much more heavily because interest on reserves and crisis preparedness made Fed deposits more attractive than private-sector alternatives. Foreign repo pool as a 'black hole' (Priority: 4/5): Foreign central banks and official institutions increasingly parked dollars at the Fed, and because those funds are Fed liabilities but not bank reserves, they effectively siphon reserves from the U.S. system. Policy fixes for the repo market (Priority: 5/5): Selgin supports a standing repo facility but argues it should be paired with lower foreign repo pool rates, caps, Treasury use of direct repo, and possibly direct access in emergencies to tame reserve volatility. Long-run path back to a corridor system (Priority: 4/5): By reducing volatile non-reserve liabilities and stabilizing reserve supply/demand, the Fed could more easily run a smaller balance sheet and potentially return to a corridor-style operating regime.

Key Arguments: The repo spike was not just a temporary market glitch; it exposed a structural shortage of bank reserves. QT removed about $600 billion in reserves, but TGA and foreign repo pool growth also removed roughly similar magnitudes of reserves from the banking system. The Treasury General Account became more attractive after 2008 because interest on reserves and crisis risk made holding cash at the Fed cheaper or safer than private alternatives. The foreign repo pool is attractive because the Fed sets its rate too high relative to alternatives, and it should be made less appealing through lower rates or caps. A standing repo facility can reduce precautionary reserve demand, especially if regulators accept Treasuries as close substitutes for reserves. Volatile non-reserve liabilities can force the Fed into an 'accidental corridor system,' causing rates to spike even when the Fed wants a floor system. The Fed and Treasury should coordinate to reduce volatility in the TGA, similar to earlier episodes when Treasury cash-management programs helped stabilize reserves. Allowing Treasury and foreign official entities limited access to a standing repo facility could provide a crisis backstop and reduce the need for a large TGA balance. If these liability-management reforms were adopted, the Fed could keep a smaller balance sheet while improving control over overnight rates. These changes could also make a future return to a true corridor system more feasible by removing the main obstacle: volatile reserve-draining liabilities.

Data Points: Repo rate spike: nearly 10% - Mid-September repo market stress reached this level while interest on reserves was around 2%. Interest on reserves: around 2% - Benchmark rate cited during the September repo-market disruption. Reserves reduced by QT: about $600 billion - Quantitative tightening from 2017 to 2019 shrank reserves by this amount. Current reserves: a little over $1.5 trillion - Approximate reserve stock in the banking system at the time of the discussion. Hypothetical reserves without QT/TGA/FRP growth: around $2.6 trillion - Back-of-the-envelope estimate if QT had not occurred and TGA/foreign repo pool had remained near pre-2008 levels. Treasury desired TGA balance pre-2008: about $5 billion - Treasury kept its Fed cash balance very small before the crisis. Treasury desired TGA balance in 2015: at least $150 billion - Treasury increased its target balance at the Fed to cover roughly a week of expenses during a future crisis. TGA balance today: roughly $350 billion - Approximate size mentioned during the episode. Proposed foreign repo pool rate adjustment: about 10 basis points below SOFR - Selgin suggests making the foreign repo pool less attractive than private-market alternatives. End-of-year reserve injection estimate: about $300 billion plus - Discussion of how much liquidity the Fed might add by year-end through repos and purchases.

Pivotal Quotes: "the black hole of the money market" — David Beckworth citing Zoltan Pozsar: Describing the foreign repo pool’s tendency to absorb reserves out of the banking system. "it would be very busy, as we know. And still, they have failed in a number of remarkable instances to get rates to behave the way they want them to" — David Beckworth: Critique of the floor system’s supposed simplicity and its repeated failures to control rates. "it’s an accidental corridor system" — David Beckworth: Characterizing the Fed’s unplanned return toward corridor-like rate behavior when reserves become scarce.

Implications: The Fed may need more than emergency repos; it must manage TGA and foreign repo pool behavior to stabilize reserves, protect overnight funding markets, and avoid repeated year-end rate spikes. These reforms could also shrink the Fed’s balance sheet and improve rate control.