BPI Symposium on Money Markets

Last week, the Bank Policy Institute hosted a symposium in New York City to discuss repo market functioning, the discount window and demand for reserves and the implication for Fed balance sheet policy. The symposium was attended by market participants, academics and the official sector.

Treasury Repo Market Functioning

Participants discussed the implications of periodic strains in the Treasury repo market, notably the spike in repo rates on Sept. 30. Most participants viewed the spikes as symptoms of limits on intermediation capacity of large bank-affiliated broker-dealers, which leave them unable to fully accommodate demands for repo market intermediation, notably at quarter ends and on settlement dates for Treasury issuance. Many attributed the lack of capacity to banking regulations, including the GSIB capital surcharges and the supplementary leverage ratio, which discourage dealers from allocating capital to low-risk, low-return business lines, including intermediating in the Treasury cash and repo markets. Others noted that, even in the absence of regulatory disincentives, the economics of allocating capital to activities that are only episodically profitable are poor. In addition to changes to banking regulations, participants identified and discussed two sets of measures that could enhance intermediation capacity in the Treasury repo market: (1) broader central clearing and (2) modifications to the Federal Reserve's standing repo facility (SRF).

The SEC has mandated that most Treasury cash and repo transactions be centrally cleared, with implementation dates of year-end 2025 for cash transactions and June 2026 for repos. Central clearing can increase the balance sheet capacity of dealers by allowing them to net some repos against reverse repos under accounting rules and for the purpose of calculating leverage ratios. Central clearing is especially promising as a means of preserving intermediation capacity under stress, when market participants may be reluctant to execute non-centrally cleared repos because of concerns about counterparty credit risk. June 2026 is a very demanding deadline, but market participants are working hard to meet it and making progress. Among the challenges is designing and implementing a functional "done-away" clearing model - that is, a clearing model that allows buy-side firms to execute repos with a wide range of counterparties but have a third-party agent clear the transactions rather than having the clearing "done with" each of the counterparties to the trades.

Participants discussed why the existence of the Fed's SRF didn't prevent the spike in repo rates at the end of the third quarter. Some pointed to what they regarded as flaws in the design of the SRF, including that SRF repos are not conducted and settled until the afternoon and are not centrally cleared and that the Fed's eligibility criteria prevent most market participants from accessing the facility. Others argued that use of the SRF entails stigma that is sufficiently severe that many of those with access remain reluctant to borrow unless alternatives are very costly. (The next section provides further discussion of the sources of stigma and measures that could make the SRF a more effective repo market backstop.)

Fed Lending Facilities and Demand for Reserves

The discussion began with a preview of an ongoing BPI survey of treasurers (results were subsequently published). The survey showed that the Fed has been having some success in encouraging banks to be more willing to point to the discount window and the standing repo facility in their liquidity planning. The most significant reason cited by banks that have become more willing was the FAQ the Fed issued this summer that stated that banks could point to the discount window, standing repo facility or FHLB advances in their internal liquidity stress tests (ILSTs) as the means by which they would monetize their liquid assets (for a discussion of the FAQ, see "A Helpful Federal Reserve Board Statement on Bank Liquidity"). Responses to other questions on the FAQ, however, demonstrated that banks still have widely varying interpretations of the rules for banks' required ILSTs, possibly reflecting caveats in the FAQ language or inconsistent guidance from their examiners.

Participants noted many different sources of stigma associated with borrowing from the Fed. Several bankers provided examples of how examiners looked askance at planning to borrow from the Fed, including by limiting such planned borrowing in ILSTs and resolution plans. Moreover, while Fed examiners are signaling a more positive view of borrowing from the Fed, examiners from other agencies viewed the discount window and SRF negatively. Stigma was also driven in part by investors' and ratings agencies' views of borrowing plus concerns about Fed disclosures of borrowing. The kludginess of discount window operations was also noted. One symposium participant observed that he had thought borrowing from the discount window must be like going to the principal's office but now realizes it is like going to the DMV.

Symposium participants suggested several ways to increase banks' willingness to borrow from the Fed. Several suggested that SRF loans should be centrally cleared to reduce balance sheet cost and that the set of SRF counterparties should be expanded to include smaller broker-dealers and hedge funds. In addition, the Fed should allow counterparties to request and receive SRF loans several times a day, not just at 1:30 p.m. Symposium participants highlighted the high cost of intraday overdrafts and liquidity shortfalls, which can result in sudden funding market pressures. Some noted that intraday liquidity could be supported by new private forms of intraday repo or by adding SRF operations earlier in the day. Many noted that the most promising way to reduce stigma was to make Fed lending more frequent. One participant suggested that the Fed begin to conduct frequent normal repo operations to help reduce the stigma associated with the SRF. Another observed that the Fed could offer primary credit through auctions of fixed amounts of term loans - similar to the auctions of discount window credit that occurred in the GFC - along with providing it through the regular discount window. Importantly, such auctioned loans would need to also be called "primary credit."

Bankers in attendance stated that their demand for reserve balances had remained unchanged or become somewhat higher. They pointed to upwardly revised estimates of how rapidly deposits could be withdrawn and pressure from supervisors as two reasons for elevated reserve demand.

Lessons for QT and the Fed's Implementation Framework

In the final section, participants discussed the implications of recent money market developments for Fed balance sheet management. As the Fed continues reducing its assets through QT, its liabilities, including reserve balances, also decline. The Fed has indicated that it will end QT when it is approaching banks' current structural demand for reserves. Many participants observed that recent indications of fragility in the repo market were the result of problems with repo market intermediation rather than a signal that the supply of reserves was getting tight.

The group discussed the implications of the debt ceiling for reserve supply and reserve management. When the Treasury runs down its deposit at FRBNY (the "Treasury General Account" or "TGA") to avoid the debt ceiling, reserve balances will rise. When the limit is raised or suspended, Treasury will refill the TGA, potentially causing a disruptive decline in reserves unless the Fed takes offsetting actions using temporary repo operations.

There was a range of views on the costs and value of volatility in repo rates. Several participants argued that the Fed should focus on keeping volatility low, in part by maintaining a larger quantity of reserves through an earlier end to QT, by keeping the ON RRP rate relatively close to the IORB rate and being willing to maintain a positive balance in the ON RRP facility as a source of liquidity for the financial system. Others judged that volatility in repo rates was useful for encouraging market participants to rebuild money market arbitrage functions and for discouraging excessive leverage at financial institutions.

Participants compared and contrasted the Fed's tools for gauging when reserve supply is getting tight, discussing the perspectives expressed by Lorie Logan, President of the Dallas Fed, and Roberto Perli, Logan's successor as the System Open Market Account (SOMA) Manager. Logan's approach focuses on repo rates being roughly equal to the interest rate the Fed pays on reserve balances as an indication that liquid assets are distributed efficiently across the system. Perli focuses on banks' demand for liquidity and the ability of the Fed to control the federal funds rate. Several participants judged that the Fed was paying too much attention to the fed funds rate rather than the repo rate. In addition, participants argued that the Fed should publish all its metrics for reserve supply tightness. Overall, participants expected QT to be brought to a close in the first half of next year.