Statements & Remarks

Remarks by Under Secretary for Domestic Finance Nellie Liang at the European Central Bank - Federal Reserve Bank of New York Workshop on Nonbank Financial Institutions

As Prepared for Delivery

Thank you for the opportunity to speak at this conference today.[1] It’s a terrific program on timely questions, and I am delighted to be part of it.

Since the global financial crisis (GFC), the risks of nonbank financial intermediaries (NBFIs) and what to do about them have become increasingly important. In the United States, while credit provided by NBFIs has been growing for decades, it started to grow more quickly than bank credit after 2012 and is now sizable, exceeding on-balance sheet bank credit.[2]

In addition, the realization that microprudential regulation and investor protection are not sufficient to ensure financial stability and prevent another financial crisis has raised critical questions about how to better regulate NBFIs. Moreover, regulations are complicated by the fact that NBFIs are very heterogenous, and spillovers from them, when they are stressed, to the rest of the financial system also are heterogeneous. 

I will offer some comments from the perspective of how to ensure NBFIs have access to liquidity in periods of high stress so that the official sector does not need to resort to extraordinary interventions to support the financial system and economic activity. In periods of stress, most NBFIs will need access to liquidity to continue lending or to facilitate a smooth failure or closure. Many NBFIs provide similar services as banks and are exposed to some of the same risks as banks, and loss of liquidity could lead to forced asset sales and can disrupt other NBFIs, banks, markets, and the broader financial system. 

NBFI access to liquidity in stress periods can come from commercial banks or it can come from markets, specifically Treasury markets. However, when banks and markets cannot or will not provide liquidity to NBFIs, as we have seen on several occasions, the official sector may feel compelled to respond. These responses are costly, as they may conflict with price stability goals of central banks and create distortions in financial market participants’ behavior, which gives rise to the need for strong ex-ante macroprudential policies. 

Acharya, Cetorelli, and Tuckman (2024) highlight the centrality of banks as liquidity providers, given their unique deposit franchise and access to the lender of last resort.[3] They provide a new systematic analysis of flows between banks and nonbanks, and a number of case studies, such as REITs, documenting the drawdowns of bank credit and liquidity commitments during the pandemic. 

This work is consistent with earlier work by the Financial Stability Board (FSB) (2013) that grew out of the attention to links between banks and “shadow banks” following the GFC, and the FSB’s current work on NBFIs.[4] It is also consistent with the growth in the deposits-to-GDP ratio, especially relative to the ratio of commercial and industrial loans to GDP over the last 40 years, as I highlighted in a recent speech.[5] In many ways these trends underscore the role of banks as liquidity providers through both the asset and the liability sides of their portfolios, as formulated in Kashyap, Rajan, and Stein (2002).[6]     

I want to add to the discussion today that NBFIs, and in particular investment funds, also rely on the Treasury market for liquidity in periods of stress. That is, like banks, Treasury markets are central to the stability of the financial system. For example, investment funds are required by the Securities and Exchange Commission (SEC) to have a liquidity risk management program that considers the liquidity risk of the funds’ holdings. In practice this means that many funds hold cash and Treasury securities for liquidity. 

But we have seen episodes in the past few years of a sharp rise in transaction costs in Treasury markets. This was evident in March 2020 at the onset of the pandemic, shown on slide 3 in the figure to the left, as well as in March 2023, when there were a number of bank runs. 

Generally, higher transaction costs are related to higher rate volatility, which is illustrated in the right panel on this slide by the upward slope to the relationship between rate volatility and transactions costs for the 10-year Treasury note. Also, transaction costs were higher than might be expected given the volatility in March 2020, as can be seen by the orange dots. These dots illustrate the “dash for cash,” when demand by investors, including investment funds, to sell Treasury securities surged and exceeded the ability or willingness of dealers to supply liquidity. 

In this episode, the Federal Reserve stepped in to purchase Treasury securities, in significant amounts for multiple months. In addition, the Federal Reserve created emergency liquidity facilities to purchase corporate bonds, which also eased pressures on the Treasury market and helped corporate bond market liquidity. More recently, during the bank deposit runs in 2023, shown by the purple dots, higher transaction costs were more consistent with the higher rate volatility. 

We have been working for the past three years to improve the resilience of Treasury market liquidity in stress periods, mainly coordinated through the Inter-agency Working Group on Treasury Market Surveillance (IAWG). This work has been structured as five workstreams: four focus on improving the resilience of the supply of Treasury market liquidity in stress periods, and a fifth focuses on reducing surges in demand for liquidity in stress periods.

We have made significant progress on the first four, to strengthen the resilience of the supply of liquidity. 

  • To improve the resilience of market intermediation, the Federal Reserve put in place the Standing Repo Facility (SRF) and the Foreign and International Monetary Authorities repo facility. In addition, earlier this month, Treasury started a limited buyback program that will help support liquidity in off-the-run securities. 
  • To improve data quality and availability, we have significantly expanded disclosures. Most recently, transaction-level data on price and volume (with caps) for on-the-run securities are being disclosed daily through TRACE. In addition, the Office of Financial Research has finalized its rule to collect data on non-centrally cleared bilateral repo transactions starting in December, which will allow the official sector to better monitor risks in the repo market. 
  • On the regulatory side, the SEC has adopted final rules to require central clearing of Treasury securities by year-end 2025 and repo by the end of June 2026, which will reduce counterparty risk and increase transparency in these markets. It also finalized rules around the dealer definition, though it has not yet finalized its proposed Regulation ATS to enhance the resilience of electronic trading venues. 

While these efforts leave us headed in the direction of a safer and more liquid Treasury market, there is more unfinished business in the fifth workstream, to reduce the potential for surges in demand for Treasury market liquidity. 

First, the liquidity mismatch in open-end bond and loan funds should be reduced, especially because its potential effects have gotten more significant as the amount of assets in these funds has risen sharply since the GFC. The SEC has proposed a rule with some options for anti-dilution tools to reduce the first-mover advantage arising from the mismatch, but there has been substantial industry pushback and the rule has not been finalized. In particular, the industry has pointed to high costs of adjusting their operations to make swing pricing work because a fund’s net asset value must be set before flows are recorded each day. This operational constraint appears to be a problem in the United States, but not in other jurisdictions which already feature swing pricing. The SEC and industry should continue to explore potential options.

Second, leverage in the Treasury market could become excessive and destabilizing. One source of such leverage is the basis trade. A basis trade opens when asset managers take long positions in Treasury futures, paying a premium, and hedge funds take short positions, which they hedge with long positions in cash Treasuries that are financed via repo. As shown in the figure on slide 5, both futures positions were growing before March 2020, and both have grown again recently. 

While the role of hedge funds in the basis trade has gotten a lot of attention because their repo transactions can be highly levered, attention now also is being paid to the demand side for futures coming from asset managers. Barth et al. (2024), for example, provide evidence that these Treasury futures positions are largely attributable to mutual funds using futures rather than Treasuries to track the duration of a benchmark bond index with Treasuries.[7] These long futures positions create leverage in the Treasury market, alongside the leverage from the hedge funds that take the corresponding short positions in Treasury futures, hedged with Treasury securities financed with repo. 

The basis trade can provide benefits by increasing Treasury market liquidity, improving integration between cash and derivatives markets, and translating demand for Treasury futures into demand for Treasury securities. But the basis trade also raises the costs of Treasury exposure in open-end funds and lengthens transaction chains for the demand for Treasury securities. 

There are two issues to consider for policy purposes. First, while bond funds may prefer futures to cash Treasuries because of higher liquidity in futures, there also are non-economic incentives. As described by the Treasury Borrowing Advisory Committee at its January 2024 meeting, there are accounting differences that require the reporting of repo borrowing as an expense (but do not apply to implied financing costs for futures).[8] This expense may raise a fund’s reported expense ratio, which is a metric often used by investors to rank a fund’s attractiveness. In addition, SEC rules and guidance interpreting the Investment Company Act of 1940 have discouraged repo relative to futures, though the SEC finalized Rule 18f-4 in 2020 to update the regulatory framework for the use of derivatives and permit open-end funds to treat repo as a derivative like futures.

Second, hedge funds in bilateral repo transactions may be getting very low haircuts on Treasury repo transactions, though some counterparties may use portfolio or cross-margining to help manage this risk. Policymakers want to prevent leverage from becoming excessive to avoid a destabilizing unwind of positions in a stress period when volatility increases and margins are raised. 

To summarize, NBFIs rely on Treasury markets for liquidity, but at the same time some NBFI practices may create vulnerabilities in the Treasury markets. These sources of potential systemic risk are currently a high priority for policymakers to address. 

While many changes have been implemented, as I outlined earlier, there is still a need for additional reforms to strengthen Treasury market resilience. To improve the elasticity of the supply of market liquidity, the banking regulators should consider changes to the supplementary leverage ratio (SLR) by excluding reserves and making the enhanced-SLR buffer countercyclical to be released in periods of market-wide stress. This change could be accomplished without reducing the overall amount of capital in the system given the recent and proposed increases in capital requirements and would allow bank dealers to be better able to make markets in periods of stress. To reduce surges in demand for liquidity, as already highlighted, reforms to reduce the liquidity mismatch and the “phantom” liquidity of open-end bond and loan funds are needed. One alternative to swing pricing that could be considered could be to lengthen the period in which investors should be able to redeem shares, from daily to the number of days it would typically be expected to take to sell the underlying securities even in normal periods. Finally, for hedge fund leverage, policymakers should assess if new data collections on bilateral repo transactions this year and the central clearing of repo with more transparent margins in two years, in addition to ongoing bank supervisory work on counterparty risk management, are sufficient to prevent excessive leverage of hedge funds.

The slides and figures referenced in these remarks are available here.


[1] I would like to thank Burcu Duygan-Bump and Brian Smith for assistance in preparing these remarks, and Shantanu Banerjee for help with the figures.

[2] Financial Stability Board, “Global monitoring report on non-bank financial intermediation,” December 18, 2023.

[3] Acharya, Viral V. and Acharya, Viral V. and Cetorelli, Nicola and Tuckman, Bruce, “Where Do Banks End and NBFIs Begin?”, March 15, 2024.

[4] Policy Framework for Strengthening Oversight and Regulation of Shadow Banking Entities - Financial Stability Board, 2013.

[5]  Remarks By Under Secretary for Domestic Finance Nellie Liang at the 2024 OCC Bank Research Symposium |, U.S. Department of the Treasury, June 24, 2024.

[6] Kashyap, Anil K, Raghuram Rajan, and Jeremy C Stein,  “Banks as Liquidity Providers: An Explanation for the Co-Existence of Lending and Deposit-Taking,” Journal of Finance, Volume 57, Issue 1, Feb. 2002.

[7] Barth, Daniel and Kahn, R. Jay and Monin, Phillip and Sokolinskiy, Oleg, “Reaching for Duration and Leverage in the Treasury Market,” May 3, 2024.

[8] See TBACCharge1Q12024.pdf (, including reference to FINRA’s clarification D.1.2 in 2019 to exclude the interest expense of repo from published expense ratios.