The Bank Wants Gilt Repo Clearing to Free Dealer Capacity. Margin Could Take It Back.

The Bank of England wants a redesign of the gilt repo market to do two jobs at once. More central clearing should let dealers net more exposures and preserve room to finance investors during a shock. More risk-sensitive margining should stop cheap leverage from becoming so large that its unwind overwhelms that room.

That is a more demanding test than making market plumbing safer. In a 1 October speech, financial stability executive Nathanael Benjamin said clearing access, leverage pricing and collateral efficiency should be treated as complements. The Bank plans a comprehensive update, including possible policy proposals, in early 2027.

The package succeeds only if the capacity released by netting remains usable after central counterparties have called margin and default resources. The public evidence shows a potentially meaningful balance-sheet gain. It does not yet show the net liquidity gain under today's positions.

The current footprint is large, but the measures differ

Hedge funds' net gilt repo borrowing stood at about £85 billion on 25 June 2026, according to Sterling Money Market Data. That is a net position across the non-bank sectors classified in the Bank's reporting, not the gross size of the basis trade. The Bank says the classification and dataset coverage remain under review.

A separate estimate says globally active hedge funds accounted for as much as 60% of secondary cash-gilt volume. Its boundary matters: it covers Tradeweb's electronic platform in January and February 2026, rather than every venue or the full year.

The often-paired £5 billion stress figure has a different date again. The Bank's first system-wide exploratory scenario used positions at 31 October 2023 and responses from about 50 firms. A later sensitivity applied a similar rates shock to hedge funds' roughly £50 billion net borrowing as of 14 November 2024. With a simultaneous 10% reduction in bank repo lending, it produced up to £5 billion of additional gilt sales, depending on funds' use of cash buffers. If banks' appetite stayed fixed, those sales would have exhausted the remaining market-making capacity.

That is not a replay at the June 2026 £85 billion footprint. It is evidence of a transmission channel, not a current loss forecast.

The clearing gain comes from a 2020 counterfactual

The capacity case rests on a 2023 Bank staff working paper using transaction-level regulatory data. For the dash for cash, it compared average daily gilt repo from 9 to 18 March 2020 with the prevailing clearing structure. Comprehensive clearing would have increased nettable repo from £472 billion to £553 billion, a gain of £81 billion or 17%.

The dealer balance-sheet calculation used a smaller sample with the necessary regulatory filings. Those firms represented about 75% of repo stock during the stress. Comprehensive clearing reduced the leverage-exposure impact of their gilt repo books by about 40%. The detailed chart reports a 3.8-basis-point aggregate leverage-ratio gain on the 13 March peak day, while the paper's abstract and later Bank summaries round it to 3 basis points.

The maturity experiment was separate. It moved non-overnight repos already expiring in the same week onto the same day. That added about £40 billion of nettable repo, equal to 4% of outstanding repo, beyond comprehensive clearing alone. It cut exposures by a further 20 percentage points and added 2.0 basis points, taking the estimated total reduction to about 60% and the detailed leverage-ratio gain to 5.8 basis points. Later summaries round the combined gain to 5 basis points.

These are material estimates, not a current replay. Respondents to the Bank's 2026 feedback statement argued that maturity transformation and existing interdealer clearing limit further netting. They supplied no quantified estimate and did not establish whether their view applied in calm or stressed markets. The Bank has not published an equivalent calculation using the 2026 book.

Zero haircuts are not one thing

Half of bilateral gilt repo haircuts are routinely set at zero, with zero common in hedge-fund transactions. Some of that can be prudent. Dealers routinely net repo against reverse repo, and portfolio margining can recognise genuine offsets.

But the Bank says formal cross-product margining appears materially available only from a few large dealers to parts of the market. A 2023 PRA review also found competitive pressure influencing near-zero haircuts. A zero applied because collateral elsewhere already covers the risk is different from a zero offered to keep a client.

That distinction makes a flat haircut floor a blunt instrument. Benjamin's speech instead points towards a portfolio approach that recognises defensible offsets while charging for concentrated leverage. Cross-margining could reduce unnecessary collateral, but only where correlations remain reliable in stress and legal and operational arrangements can withstand a default.

Capacity can move rather than appear

Central clearing replaces bilateral exposures with other claims on liquidity. Under the Bank's 2025 discussion paper, LCH recalibrates initial margin at least daily against 99.7% of historical five-day moves and can add charges for concentration and credit risk. Members also contribute to a mutualised fund sized for the simultaneous default of the two largest exposures. Variation margin is exchanged daily in cash.

Sponsored access can leave a bank funding a client's margin and default-fund contribution, or guaranteeing losses beyond the client's prefunded resources. Wider clearing can therefore shrink a leverage-ratio denominator while increasing the cash and contingent resources demanded of the same intermediaries.

The scale of that offset has not been published for gilt repo. Broader UK evidence shows the issue: during March 2020, initial margin at UK clearing houses rose 31% to £58 billion, while average daily variation-margin calls were five times their January and February level. Those figures cover UK CCPs generally, so they cannot be subtracted directly from the repo netting estimate.

Maturity standardisation adds another unresolved trade-off. By construction, moving weekly expiries onto one day increases netting. It also concentrates refinancing dates. The Bank has quantified the first effect, but not the same-day liquidity demand under a current stress.

The right scorecard for early 2027 is therefore not gross clearing adoption. It is dealer exposure released, less cash margin, default resources and refinancing liquidity required under the same scenario. Until the Bank publishes that matched test on current positions, clearing is a credible route to more capacity, not proof that the constraint has disappeared.