Improving Capacity and Resiliency in US Treasury Markets: Part II

Proposals for Reforming US Treasury Markets

In Part I of this blog series, we provided an overview of the important role that the U.S. Treasury markets play in the financial system and in the transmission of monetary policy. We also provided a brief history of recent disruption events that have led to growing calls for reform of the U.S. Treasury markets.

In Part II we discuss a variety of ideas for reforming the markets.

The market disruption events we discussed in our previous blog – the 2014 “Flash Rally,” the 2019 stresses in the Treasury-backed repurchase (“repo”) market, and the 2020 COVID crisis – have led to growing calls for reform of the U.S. Treasury markets. As we noted, the Financial Stability Board (“FSB”) and other regulators are looking at these issues, and the topic has been discussed at length in many white papers and academic articles. For our purposes, we will focus on two papers that have attracted particular attention: a paper by Stanford University Professor Darrell Duffie[1], and one produced by former senior Federal Reserve staffers Nellie Liang (also of the Brookings Institution and now the U.S. Treasury Department) and Pat Parkinson.[2]

Although by no means exhaustive of the ideas circulating in public debate around Treasury market reform[3], it does seem that there are four broad categories of reform currently being debated. These are (1) greater use of centralized clearing and “all-to-all” trading platforms; (2) the creation of a permanent standing re-purchase (“repo”) facility (SRF); (3) changes to banking regulation, particularly for bank holding companies; and (4) improved data and disclosure by market participants. While each of these approaches have been discussed on a standalone basis and most had been suggested prior to the March/April 2020 events, Liang and Parkinson argue that these reforms ought to be considered holistically as a single package. We summarize and then discuss each of these in brief below.

Summary of Proposals and Possible Costs and Benefits

Expanded Centralized Clearing

Summary:

U.S. Treasury cash transactions would be required to be cleared via a CCP (would also apply to repo transactions, but these are increasingly cleared centrally)

Costs/Benefits:

  • May increase market capacity by reducing demands on dealer balance sheets through netting efficiency gains and possible greater use of “all-to-all” platforms
  • However, could increase trading costs and create concentration risk at the CCP
  • Impact on past stresses: greater central clearing in repo markets may have contributed to some stresses in those markets in 2019 by bringing new actors into markets. Central clearing unlikely on its own to have had a major impact in mitigating the 2020 dash-for-cash

Standing Repo Facility

Summary:

Federal Reserve would create an ex-ante standing repo facility (“SRF”) that would offer collateralized funding to a broad range of dealers

Costs/Benefits:

  • Could ensure a backstop  source of market liquidity, and may encourage greater dealer participation in the markets under both normal and stressed conditions
  • The rate charged by the SRF will be crucial to its success; will need to avoid pricing too high to stigmatize usage or too low, which would create over-reliance and distort the market
  • Breadth of access to the SRF, political questions over whether Fed ought to be a “market maker of last resort” and what role the Fed should play in the market would need to be resolved
  • Impact on past stresses: likely would have helped in 2019 repo event. Unclear if SRF would have helped without addition of more active Fed market interventions

Expanded Dealer Balance Sheet Capacity/Leverage Ratio Reform

Summary:

Expand dealer balance sheet capacity by permanently excluding Treasuries and reserves from the Supplementary Leverage Ratio (“SLR”)

Costs/Benefits:

  • SLR as currently designed constrains ability of dealers to absorb increased Treasury inventories, particularly at a time when deposits are also increasing
  • Impact on past stresses: temporary relief appeared to have positive effect on dealer capacity during 2020 event, though effect as a standalone measure likely would have been limited

Enhanced Reporting and Disclosure

Summary:

Require greater disclosures from market participants, particularly in bilateral repo market and from funds

Costs/Benefits:

  • Added disclosures and data reporting after 2014 Flash Rally have generally proved beneficial
  • New reporting mechanisms under discussion would generally not provide real-time market data and therefore would be of limited use during a crisis, though they could help us better understand the crisis after it occurred
  • These data collections would also likely struggle to measure liquidity in the less frequently traded off-the-run securities markets

Expanding Centralized Clearing

The idea of expanding or mandating central clearing in the Treasury markets has been under discussion for some time, with several government and non-government reports examining the idea in recent years.[4]   In his paper, Duffie proposes a “study of the costs and benefits of mandating the central clearing of Treasury transactions of all firms that are active in the market.” In this potential scenario, a central counterparty (“CCP”) would intermediate all trades in the market, whether those are interdealer trades or dealer-to-client trades. Indeed, much of the trading in the Treasury repo market is already centrally cleared via the Fixed Income Clearing Corporation (“FICC”), which is a CCP.

Duffie argues that a broad central clearing mandate ought to be considered because “the size of the Treasury markets will outstrip the capacity of dealers to intermediate the market.” More specifically, central clearing would free up “the amount of dealer-balance sheet space necessary to maintain liquid markets,” since capital and leverage requirements for dealers would be reduced because of multilateral netting of trades at the CCP.  Centralized clearing could also facilitate a greater movement toward “all-to-all” platforms for trading securities, which could increase balance sheet capacity for dealers by disintermediating them from some trades i.e., providing for direct trading between non-dealer buyers and non-dealer sellers (this would be unlikely to disintermediate dealers entirely, since their liquidity provision would continue to be needed, particularly during stressed periods). Beyond increasing capacity, centralized clearing may also have other ancillary benefits, such as increasing the transparency of settlement risk and improving market safety by lowering exposure to settlement fails.

However, others have raised concerns about the move to central clearing and the need for additional clearing in different parts of the market, most notably the repo market. FICC clearing in the repo market has been enhanced by the broadening out of FICC’s sponsored program which has allowed FICC members (i.e., dealers) to facilitate their sponsored clients’ trading activity and act as processing agents on their behalf for all operational functions, including trade submission and settlement with the CCP.[5] While this has certainly expanded the cleared market for Treasury repo transactions, there is evidence that centralized clearing has facilitated greater repo lending from money market funds (“MMFs”) to leveraged firms such as hedge funds, a trend that may have contributed to the market stresses experienced in September 2019.[6]

There are other factors that will need to be studied before moving to a centralized clearing model.  For example, central clearing could significantly increase the cost of trading in the cash Treasury markets.[7] There is also the concern that centralizing trading into a CCP would concentrate too much settlement risk into one institution, meaning that any operational or liquidity stresses that occur at the CCP could translate into widespread illiquidity events in the Treasury markets. Then there is the bigger question of whether greater central clearing would have solved for any of the most recent stresses that have occurred in the markets, particularly the March 2020 COVID event. Even with most Treasury trades being centrally cleared, it is highly unlikely that sufficient capacity would have been freed up to absorb the “dash-for-cash” by investors that occurred last year.

Mandatory central clearing will certainly be a key element of the debate around Treasury market reform. However, as both Duffie and Liang and Parkinson note, it should be subject to thorough study to assess both its costs and benefits—and particularly how different segments of the market may be impacted by additional clearing– before reforms are undertaken. That in turn will likely require additional data on the markets (see below), as well as a more thorough understanding of the events of last March in particular.

Creation of a Standing Repo Facility

Ultimately, the market turmoil that we witnessed last year was stemmed by a series of extraordinary central bank interventions across the world to provide liquidity to markets. In this sense, as some have observed, central banks took on a new role not just as a “lender of last resort” (ensuring bank liquidity) but also as “market maker of last resort” (acting as a backstop for market liquidity).[8] This has raised the question of whether a standing market liquidity facility, much like the discount window for banks, should be created to offer secured financing to a broader range of market participants.

For our purposes, we will discuss the slightly narrower concept of a standing repo facility (SRF) proposed by Liang and Parkinson.[9] This proposal would create an ex-ante facility open to all broker-dealers, both those affiliated with a bank and those that are not affiliated with a bank. The facility would operate much like the temporary Primary Dealer Credit Facility (PDCF) created in March of last year: it would offer collateralized funding (possibly limited to Treasury and agency securities) to dealers in sound financial condition at a rate of interest designed to be “high enough to ensure that the facility is used only in abnormal market conditions but not so high as to stigmatize use of the facility and thereby defeat its purpose.”[10] Unlike the PDCF, it would be open to a broad range of dealers, encouraging smaller dealers to engage in intermediation in the Treasury markets under both normal and stressed conditions. This would, according to Liang and Parkinson, add depth and diversity to the market.

However, access to a guaranteed liquidity backstop could create a moral hazard problem, encouraging firms – particularly firms that are not subject to prudential oversight – to take on excessive leverage or maintain insufficient liquidity reserves. It could also exacerbate an un-level playing field between more heavily regulated bank broker-dealers and non-bank dealers. As a result, Liang and Parkinson argue that any standing repo facility participant would need to “meet prudential requirements established by the Federal Reserve, in consultation with the Securities and Exchange Commission (SEC).”[11] While this sounds straightforward in theory, in practice many non-bank dealers may view prudential regulation as too steep a price to pay to participate in the SRF, thus attenuating its positive benefits.

Success of any SRF would also be highly contingent on the rate charged. If the price is too close to market rates, it could distort the market, and further enlarge the Fed’s role in the market on a day-to-day basis. If it is set too high, then few market participants will use the facility as it would recreate the stigma that has become associated with discount window lending. Other details, such as the type of collateral that will be accepted, the trading hours during which the SRF will be available, and how SRF transactions will be cleared, will all be crucial to the success of any facility. Of course, there is also the question of how prudential rules for non-banks would be determined, and the jurisdictional issues this would raise between the Federal Reserve and the SEC.

While all of these critical details would have to be worked out before a standing repo facility could be created, there is a more fundamental political question that will need to be answered first: are policymakers comfortable with the Federal Reserve serving as a permanent guarantor of market liquidity, with the attendant moral hazard problems that could be created? That is a debate that has yet-to-be-settled.

Increasing Dealer Balance Sheet Capacity – Leverage Requirements

One of the most important constraints on dealer capacity stems from regulatory requirements, specifically the Supplementary Leverage Ratio (SLR). A post-financial crisis reform, the SLR (like all leverage ratios) is a measure of total capital to total assets, though unlike other leverage rules it also includes certain off-balance sheet exposures in its denominator. The SLR and other leverage requirements are intended to be risk-agnostic backstops to risk-based capital requirements. Under risk-based capital rules, Treasuries and reserves held at Federal Reserve banks are assigned zero percent risk-weights, essentially treating them as equivalent to cash (i.e., free of credit risk). Under the SLR and other leverage ratios, those holdings receive equal weight to far riskier and higher-returning assets.

This difference in treatment can turn the SLR into a binding constraint for some large dealer banks, placing constraints on their intermediation capacity even during normal times. For example, as Liang and Parkinson note, the share of gross inventory positions of Treasury securities held by dealers (most of them bank-affiliated) has not changed much since 2008, despite a dramatic increase in the inventory of Treasuries being issued.[12] However, during periods of market stress, those capacity constraints can contribute to liquidity shortages in the Treasury markets, as seen in March of last year. Bank dealers simply did not have the balance sheet capacity to purchase the volume of Treasuries that were being sold by funds, PTFs, institutional investors, and foreign governments, in part because doing so would cause them to breach their SLR capital requirements (as discussed in a recent SIFMA blog, these balance sheet capacity constraints are further compounded by the increase in bank deposits that occurs as businesses and investors liquidate investments, and as the government engages in economic stimulus).[13]

For this reason, the Federal Reserve and the other banking agencies issued temporary relief in late March of last year exempting both Treasuries and reserves held at Federal Reserve banks from the SLR, both considering the then-market conditions and the forthcoming growth of the Federal Reserve’s balance sheet and expected high volume of Treasuries issuance. Many in the industry, including SIFMA, have long argued that these changes to the SLR should be made permanent, while others (including Liang and Parkinson) have suggested that the exemption only be extended to reserves and/or that regulators consider offsetting any exemptions by an increase in the overall SLR. The Federal Reserve recently announced that it would let these temporary changes to the SLR expire but indicated it would invite public comment on permanent SLR modifications “to ensure it remains effective in an environment of higher reserves.” The outcome of that process will likely help determine the success of any broader effort to reform the Treasury markets.

Improved Data Collection and Disclosure

One common theme in recent papers and reports is the need to gather more data, particularly data related to non-bank dealers and leveraged funds (though there is widespread acknowledgment that data has improved since the 2014 Flash Rally event). Liang and Parkinson highlight, for example, the need for additional data on the bilateral repo market. They also call for greater disclosures from market participants, including greater disclosures from funds (Form PF); reporting of balance sheets and activities of non-bank dealers that participate in their proposed SRF; and disclosure of data on cleared repo, particularly repos cleared through FICC’s repo service.

While greater data in some aspects of the market may be helpful, the reporting mechanisms under discussion would generally not provide real-time market data and therefore would be of limited use during a crisis, though they could help us better understand the crisis after it occurred. These data collections would also likely struggle to measure liquidity in the less frequently traded off-the-run securities markets. Nevertheless, there is little doubt that further data on aspects of the Treasury markets will be needed before significant reform can be undertaken.

Conclusion

US Treasury Market Reform Proposals - SIFMA

Recent events highlighted longer-term structural questions in the Treasury markets, particularly the growing mismatch between the volume of Treasury issuance and the capacity of the system to intermediate the trading of those instruments. Some of the reforms discussed here appear more feasible to implement than others; for example, modifications to the SLR would be relatively straightforward, while a move to a standing repo facility may take longer to put in place owing to technical and political challenges. What is certain is that more study is needed to determine the impact of many of these proposals, both individually and collectively, on the Treasury markets. Moreover, we are likely to see other reform ideas emerge over the coming months as the FSB, U.S. Treasury, and others continue to study recent events and collect data on the markets. We will keep you updated in future blogs and SIFMA publications as this debate over Treasury market reform evolves.

Dr. Peter Ryan is a Managing Director and Head of International Capital Markets and Prudential Policy at SIFMA.

Robert Toomey is Managing Director and Associate General Counsel, Rates at SIFMA.

[1] Darrell Duffie, “Still the World’s Safe Haven? Redesigning the U.S. Treasury Market After the COVID-19 Crisis,” Hutchins Center Working Paper #62, Brookings Institution, June 2020. Available at: WP62_Duffie_v2.pdf (brookings.edu).

[2] Nellie Liang and Pat Parkinson, “Enhancing Liquidity of the U.S. Treasury Market Under Stress,” Hutchins Center Working Paper #72, Brookings Institution, December 16, 2020. Available at: WP72_Liang-Parkinson.pdf (brookings.edu).

[3] See, for example, a recently released draft paper from Andrew Metrick and Daniel K. Tarullo, “Congruent Financial Regulation,” Brookings Papers on Economic Activity, BPEA Conference Drafts, March 25, 2021. Available at: BPEASP21_Metrick-Tarullo_conf-draft.pdf (brookings.edu). We will discuss the ideas in this paper among others in future publications.

[4] See, for example, “The Joint Staff Report on the U.S. Treasury Market on October 15, 2014,” July 13, 2015. Available at: Joint_Staff_Report_Treasury_10-15-2014.pdf; U.S. Department of the Treasury, “A Financial System that Creates Economic Opportunities: Capital Markets,” October 2017. Available at: A Financial System That Creates Economic Opportunities: Capital Markets (treasury.gov); Treasury Market Practice Group (TMPG) “Best Practice Guidance on Clearing and Settlement,” July 2019. Available at: CS_BestPractices_071119.pdf (newyorkfed.org).

[5] See DTCC’s overview of the sponsored program at https://www.dtcc.com/clearing-services/ficc-gov/sponsored-membership.

[6] E.g., see see Fernando Avalos, Torsten Ehlers and Egemen Eren, “September stress in dollar repo markets: passing or structural?,” BIS Quarterly Review, December 2019. Available at: September stress in dollar repo markets: passing or structural? (bis.org).

[7] For example, see the TMPG Best Practice Guidance, p. 3.

[8] See Andrew Hauser, “From Lender of Last Resort to Market Maker of Last Resort via the Dash for Cash: Why Central Banks Need New Tools for Dealing with Market Dysfunction,” January 7, 2021. Available at: Andrew Hauser: From lender of last resort to market maker of last resort via the dash for cash – why central banks need new tools for dealing with market dysfunction (bis.org).

[9] This idea was previously proposed by Joseph Gagnon and Brian Sack in their paper “Monetary Policy with Abundant Liquidity: A New Operating Framework for the Federal Reserve.” Peterson Institute for International Economics, Policy Brief 14-4, January.

[10] Liang and Parkinson, p. 7.

[11] Ibid., p. 2.

[12] Liang and Parkinson, p. 5.

[13] The Liquidity Coverage Ratio (“LCR”) further compounds the constraint posed by the SLR.  The LCR requires the largest firms to hold on balance sheet High Quality Liquid Assets (“HQLA”) which are generally comprised of cash and U.S. Treasuries.  The size of the buffer is sized based on the potential funding needs resulting from a 30-day stress period i.e., the buffer acts as capital to offset the negative implications of a liquidity stress event.  Nonetheless, the need to maintain this buffer consumes considerable leverage ratio capacity which could be otherwise deployed to provide capacity to the markets.