Gilt edged resilience: strengthening liquidity provision in the repo market

By Sarah Breeden, Deputy Governor for Financial Stability.^{footnote [1]}

Few markets matter more to the UK financial stability and the real economy than the UK government bond, or gilt, market.

The cash gilt market not only finances government but also provides a benchmark for pricing across all sterling assets and underpins regulatory requirements.

The gilt repo market is, in my view, even more critical. It supports the liquidity and functioning of the cash gilt market, facilitates the flow of cash and collateral through the system, and helps underpin the transmission of monetary policy.

Repo markets are the core plumbing of our financial system. When they work, they support financial stability and growth. When they fail, stress can spread fast. And like all plumbing, any failure can have very unfortunate consequences. Both the March 2020 ‘Dash for Cash’ and the 2022 liability-driven investment (LDI) episode showed how structural vulnerabilities in liquidity supply, combined with surging demand, can amplify shocks to the point of requiring central bank intervention.

It is essential that we build sufficient resilience in these markets such that similar dynamics do not occur again. The importance of building such resilience is recognised internationally. Similar efforts are being taken forward in other jurisdictions, most notably the forthcoming clearing mandate for US Treasury cash and repo markets.

The discussion paper (DP) the Bank published last year, which explored the benefits of greater central clearing and the introduction of minimum haircuts on non-centrally cleared repo transactions, was an important step forward. We are immensely grateful for industry’s constructive feedback, which we published in April.

Further engagement is essential because doing nothing is not an option. These are complex reforms. And we recognise their design will be critical to achieving the intended financial stability benefits of enhancing gilt repo resilience while balancing any costs, operational considerations and market impact. Any potential changes will take time and be carefully calibrated, underlining the importance of industry feedback.

This blog addresses three of the key challenges we received to our DP and sets out how market participants can further help us ensure any reforms are best tailored to sterling markets.

The need for further resilience?

Before addressing those challenges, it is worth setting out why further resilience is needed.

Events of recent years have provided some encouraging news. As highlighted in the July 2026 Financial Stability Report, the gilt market absorbed sizeable moves in yields and activity without notable disruption following the outbreak of conflict in the Middle East. Despite significant deleveraging by hedge funds, liquidity metrics, such as bid-offer spreads, remained broadly in line with volatility – and better than during the ‘Dash for Cash’ and LDI stress episodes (Chart 1). The same was true following US announcements on trade policy in April 2025 – commonly referred to as ‘Liberation Day’. Conditions in core government bond markets deteriorated but ultimately market functioning remained orderly. We came close then, however, and conditions could have worsened absent the US announcement of a tariff pause.

This resilience in current conditions does not guarantee similar outcomes in future, however. While our first system-wide exploratory scenario (SWES) exercise demonstrated that actions taken by authorities and market participants following the LDI episode improved market resilience with money market funds and LDI funds now holding greater self-insurance, it also underscored the need for further work. In particular, it highlighted that banks would have pulled back on repo financing more aggressively in a more severe stress. It also showed how the structure of gilt markets – including firms’ positioning, levels of non-bank financial institution (NBFI) resilience and banks’ ability to intermediate – impacts resilience too.

Gilt markets have undergone significant structural changes since that exercise was conducted at the end of 2023. Globally active hedge funds have materially increased their footprint and are now estimated to account for as much as 60% of secondary market volumes in the gilt cash market.^{footnote [2]} In the gilt repo market, these hedge funds have turned from being net cash lenders to significant net cash borrowers as their trading strategies have evolved, with their net borrowing position standing at £85 billion in the most recent data (Chart 2).^{footnote [3]} This makes them more exposed to changes in financing conditions and increases the likelihood that their distress triggers disruption in gilt markets, particularly if net sales were to overwhelm dealers’ intermediation capacity. Indeed, as noted in the SWES final report, a similar scenario run now that hedge funds have increased their repo borrowing could have resulted in an additional £5 billion of gilt-sales – more than doubling the sales recorded in the SWES – fully exhausting banks’ market making capacity.

The increasing footprint of hedge funds in the gilt market is not a bad thing: quite the reverse. It has provided a counterbalance to the reduction in demand from traditional longer-term investors like pension funds. As flagged in a recent speech by Catherine Mann,^{footnote [4]} price-elastic investors such as hedge funds, have been an important source of the demand for newly issued gilts. 

However, the growing footprint of these leveraged and price-sensitive players that are active across different markets and jurisdictions brings new risks. While they can support liquidity in normal market conditions, their positioning can adjust quickly to changes in the risk environment and funding conditions, sometimes amplifying volatility. And because many operate across jurisdictions and asset classes, spillovers across markets and borders can be swift and large.

These risks underscore the importance of building further resilience. Our primary goal is to ensure liquidity provision in our core markets remains resilient in the face of market shocks. Reforms may also affect the level of leverage in the NBFI sector. But this is not our primary focus, which is on the consequences of when that leverage is poorly managed. Ongoing engagement with industry is essential to ensure any reforms are best tailored to sterling markets. The responses to the DP have provided helpful insights. With that in mind, I want to share some thoughts on three of the key challenges that were raised.

Challenge 1: The netting benefits from central clearing are limited.

Some respondents told us that netting opportunities in the gilt repo market are limited due to the maturity transformation performed by dealers.

Current market structures mean that the benefits are indeed smaller than in the US Treasury repo market, but they are still significant. Bank research shows that during the Dash for Cash, comprehensive central clearing of gilt repo would have increased the amount of nettable repo by £81 billion, reducing exposures on UK dealers’ balance sheets by 40% and increasing their average leverage ratio by 3 basis points.^{footnote [5]} That’s meaningful both at the product level and in aggregate even without any further standardisation in maturity dates, which may occur as the level of clearing increases and dealers seek to maximise balance sheet efficiencies. Such standardisation would have increased the stock of nettable repo by a further £40 billion, reducing dealers’ exposures by a further 20 percentage points and increasing their average leverage ratio by a further 2 basis points.

Those benefits should grow as market structure evolves. Experience from derivatives and US Treasury repo markets suggest clearing can itself drive change, supported by innovation in access models and contract design. SwapClear is a useful example: notional values doubled from £100 trillion to £200 trillion between the announcement and the launch of the derivatives clearing mandate, and as of 2026 they stand at roughly £500 trillion (Chart 3). During this period, the clearing offer expanded to other products and evolved, increasing the efficiency of netting sets, and new client clearing models emerged.

Any assessment of central clearing must consider how market structure may adapt over time. We are not naïve. We recognise that change takes time and will depend on the development of new clearing models and regulatory calibration. This is a transition that will take time and need managing, but early discussion is essential.

Cross-margining^{footnote [6]} was highlighted by some respondents as one way to increase the benefits of central clearing. It can unlock cross-portfolio netting benefits and reduce margin costs for both dealers and the buy-side. But it is not risk free: correlations may break down in stress and cross-margining arrangements add complexity, particularly in default management across central counterparties (CCPs). Regulators and CCPs must assess and mitigate these risks carefully.

Cross-margining – across products, currencies or CCPs – is an area we wish to explore further. We would welcome further engagement with market participants on which cross-product or cross-currency netting sets would bring the largest benefits.

We are also continuing to explore the efficacy of different access models which could broaden CCP membership and therefore increase the scope of the netting sets. In various jurisdictions, including the UK, we have seen attempts to introduce ‘sponsored models’, through which participants can become members with varying levels of ‘guarantees’ provided by sponsors. These models can help ease barriers to access but could also have consequences for financial stability as they change the distribution of risk at the CCP. We would welcome input from market participants on whether there are barriers to offering or using sponsored models that are specific to the UK context, and if there are reflections on the experiences in other jurisdictions around which models work best.

Challenge 2: Current zero or near-zero haircuts in the non-cleared market are appropriate.

Some respondents said that zero, or near-zero, haircuts in the non-cleared market reflect firms’ portfolio margining practices and that these practices were effective for managing counterparty credit risk.

To be clear, portfolio margining can be valuable and deliver meaningful efficiencies. But it must be done prudently with robust legal documentation, strong governance, and rigorous stress testing to guard against the breakdown of historical correlations

There is also evidence that commercial pressures may be contributing to the prevalence of near-zero haircuts in the non-centrally cleared gilt repo market. Half of all haircuts are routinely set at zero in the bilateral market, with haircuts on hedge fund repo transactions commonly set at zero (Chart 4). Netting across repo and reverse repo exposures is typical practice by dealers, but supervisory information suggests cross-product margining between fixed-income products appears to be only materially and formally offered by a few large dealers to certain segments of the market. That suggests some zero-haircut transactions may not be fully explained by margining practices. The 2023 Prudential Regulation Authority Fixed Income Financing Review also found evidence that near-zero haircuts may reflect competitive pressures rather than prudential risk management.

Even where portfolio margining is taking place, it is not clear that haircuts in the systemically important gilt repo market should be zero. Appropriately set haircuts can support resilient liquidity provision by reducing counterparty credit risk – a key constraint on the expansion of repo lending in the SWES. Minimum haircuts may also help reduce the relative size of procyclical increases in collateral calls in stress.

Any potential design of minimum haircuts would take account of portfolio-level approaches. This would capture the full spectrum of risks across a firm’s exposure, while also avoiding undue costs for market participants.

A well-calibrated approach need not increase overall margin costs assuming portfolio-wide collateral levels are adequate. Banks’ activity with hedge funds is broad, including both lending and borrowing securities, over-the-counter derivatives and prime brokerage services.^{footnote [7]} If margin is genuinely being set by banks on a portfolio basis, an increase in the haircut applied to one part of the book should be offset elsewhere.

We would welcome further feedback from industry on calibrating a risk-sensitive portfolio-approach to minimum haircuts. We would also welcome further insights into banks’ portfolio margining practices, including the roll of cross-product margining and competitive pressures.

Challenge 3: Greater central clearing will exacerbate liquidity pressures during periods of stress.

A number of respondents told us that the procyclical nature of CCP initial margin calls would exacerbate liquidity pressures during periods of stress.

CCP models are designed to increase initial margin requirements when markets become volatile. This is a feature, not a flaw: it keeps positions prudently collateralised and protects the wider CCP membership from avoidable loss mutualisation.

Margin procyclicality can be mitigated in centrally cleared markets. As supervisor of UK CCPs, the Bank monitors margin models to ensure there are sufficient anti-procyclicality considerations built in and has the powers to require changes where necessary.^{footnote [8]}

We recognise that clearing models involve a trade-off between participant costs, coverage and reactivity to market conditions that needs to be managed.^{footnote [9]} Lowering a margin model’s reactivity to a period of volatility, usually means higher initial margin in ‘normal’ times. We see similar trade-offs elsewhere in the regulatory framework. Indeed, the question of how much capital should be required of banks in good times – accepting that requirements could reduce lending – to ensure that losses can be absorbed without harming the economy during stress requires us to look at similar trade-offs.^{footnote [10]}

International bodies have also taken significant steps. The 2022 BCBS-CPMI-IOSCO resilience guidance outlines expectations for how CCPs should identify, assess and mitigate procyclicality in initial margin and strengthens expectations on transparency and predictability.^{footnote [11]} The risks from procyclicality can also be mitigated through the increased preparedness of clearing members and their clients. The Financial Stability Board (FSB) has set out policy recommendations to ensure non-bank financial institutions can meet large initial or variation margin calls without destabilising markets.^{footnote [12]}

So, while risks of liquidity stress from excessive margin procyclicality in centrally cleared markets are real – they can and will be mitigated.

Procyclical dynamics are also not unique to centrally cleared markets. Haircuts on non-centrally cleared gilt repo transactions are just as vulnerable. Indeed those charged on LDI funds and pension funds increased procyclicaly during the September 2022 stress episode and remained elevated for some time, largely reflecting judgements about higher credit risk (Chart 4).^{footnote [13]} Similarly in the SWES scenario, many banks said they would apply higher haircuts on new repo trades, often doubling the original haircut. The impact can be largest when margins start from zero or very low levels, as they currently are in the majority of the non-centrally cleared market.

We would welcome views from participants on whether procyclicality is greater in centrally cleared rather than bilateral markets and whether any further measures might be needed to reduce these risks.

Next steps: working towards longer-term reforms.

The resilience of the gilt repo market is fundamental to financial stability and the real economy. Further action is needed so the market can absorb rather than amplify stress.

Further engagement is essential because doing nothing is not an option. The Bank will continue to carefully consider how the reforms set out in our DP – greater central clearing and minimum haircuts – could be tailored to the specific features of sterling markets.

Any reform requires partnership. This work will reflect feedback received so far. We will explore changes in market structure that would enable the greater adoption of central clearing, and measures that would support more prudent risk management practices and margining in the non-centrally cleared gilt repo market. We will progress this work over 2026, publishing further analysis to inform the debate.

These are complex reforms that will take time. We recognise their careful design will be critical to achieving the intended financial stability benefits of enhancing gilt repo resilience while balancing any costs, operational considerations and market impact.

Continued engagement is essential to ensure we get this balance right. I encourage market participants to engage with the issues and questions set out in this post. How could market structure evolve as central clearing increases and which cross-product or cross-currency netting sets would bring the largest benefits? Are there barriers to sponsored models that are specific to the UK market? Are procyclical dynamics greater in centrally cleared rather than bilateral markets and are any further measures needed to reduce these risks? How can a risk-sensitive portfolio-approach to minimum haircuts be best calibrated?

Change of this kind will likely take years, not months. But the discussion must start now – because the prize is not just lower risk, but a stronger and more efficient market for all.

Bank Insights articles do not necessarily represent the views of the Bank of England’s policy committee members.