Government debt: the buyers who borrow

WHO HOLDS GOVERNMENT DEBT? · Part 3

They hold the bonds, borrow most of the purchase price and hedge part of the risk. How reliable is a source of demand whose strength depends on its ability to wait?

A fund can own a government bond while trying to offset most of its interest-rate exposure. The purchase is real. Much of the money comes from a lender. A separate position in the futures market is meant to counter part of the bond’s price movement. Finding the economic creditor of a government therefore requires opening more than one contract. ^[4][9]

On October 6, 2026, the IMF brought these investors back into the financial-stability debate. Hedge funds, a diverse group of alternative investment vehicles, help markets function; their borrowing and overlapping positions can also amplify stress. The same day, the industry association MFA defended their role and challenged broad restrictions on leverage. For government bond markets, that debate raises a practical question: how dependable is demand when the buyer is borrowing too? ^[2][17]

The first two parts of this series followed a bond through issuance and into the balance sheets of the banks that distribute it. This instalment reaches a buyer whose holding period depends on the economics of a trade. A long-dated government promise has entered a portfolio that can be reassessed every day.

Using futures to adjust interest-rate exposure

A presentation to the US Treasury’s advisory committee on January 30, 2024 provides a useful starting point. Its authors describe why asset managers use Treasury futures: to adjust interest-rate exposure, change portfolio allocations and simplify operations. The committee consists of market professionals. This is their account of how investors use the market. ^[1]

Consider a manager who favours corporate bonds for their additional yield but wants the portfolio’s sensitivity to interest rates to remain close to its benchmark. Buying Treasury futures can supply some of that exposure. The position moves with the value of the relevant government bonds, while initially requiring only a fraction of the reference amount in collateral. The manager has separated where the money is invested from part of the risk the portfolio carries. ^[1][6]

Another participant sells the futures. Seen in isolation, that sale might look like a wager against bonds. Yet the seller may also buy physical Treasuries. A fall in their price would then tend to be offset by a gain on the short futures position; a rise would produce the reverse pattern. The offset is imperfect and depends on the instruments and their respective sensitivities. ^[4][5]

The fund putting the two positions together connects different preferences. One investor wants synthetic exposure. Someone else must hold the securities. The fund agrees to finance and manage the bond position in return for a relative-price opportunity. Demand for interest-rate exposure can reach the physical market without the end investor arranging the bond purchase and its financing personally.

This changes the meaning of “who holds the debt?” The economic recipient of the coupon, the provider of most of the cash and the investor retaining the interest-rate exposure may occupy different places in the chain. Putting them under a single statistical label would obscure the arrangement connecting them.

A price difference that must survive until exit

The best-known version is the cash-futures basis trade. A fund buys a Treasury and sells a related futures contract, seeking a profit from their relative prices after funding costs. Under the quotation convention explained by CME, gross basis is the bond’s clean price, excluding accrued interest, less the futures price multiplied by a conversion factor. That raw number leaves several costs and contractual choices unresolved. ^[4]

The conversion factor deserves a moment. A Treasury futures contract can accept delivery of several securities with different coupons and maturities. Its rules adjust the amount paid to the seller according to the bond delivered. The fund compares prices on that adjusted basis and considers which eligible security is economically most attractive to deliver, known as the cheapest-to-deliver bond. ^[5]

Profitability also depends on coupon income, purchase and exit prices, financing and delivery choices. An option-adjusted return measure such as the one used by Federal Reserve researchers has a different meaning and sign convention from simple gross basis. Reporting that “the basis is positive” without specifying the definition can therefore leave readers comparing two quite different indicators. ^[10]

The practical idea is straightforward. The fund wants sufficient compensation to acquire, carry and eventually dispose of the bond while hedging much of the rate risk. It can exit by delivering the security against the futures contract or by closing both positions in the market. Delivery terms link the instruments’ values. Until then, their relative price can move against the fund, and financing has to remain available. ^[4][9]

The cash accounting matters. Selling futures with a $100 million reference amount does not put $100 million into the fund’s bank account at inception. It creates an exposure and settlement obligations. Gains and losses are settled over time, and an initial-margin deposit backs performance. The money for the physical bond purchase must come from another source. ^[6][7]

A hundred million in bonds, three million in capital

We can make the structure visible with a wholly hypothetical portfolio. The fund buys $100 million of bonds. A lender supplies $99 million through a repurchase agreement, or repo. Legally, the security is sold under an agreement to repurchase it. Economically, the arrangement supplies cash against an asset. Here the difference between the collateral’s value and the loan is a 1% haircut. ^[9]

The fund contributes $1 million of its own resources to complete the purchase. It posts another $1 million as initial margin for the futures hedge and keeps $1 million immediately available. Total own capital allocated to the strategy is therefore $3 million. The bond position is 33.3 times that capital. This specific ratio compares the physical position with all the strategy’s own resources, including the margin deposit and liquidity buffer.

The denominator changes the story. Counting only the $1 million that tops up the repo would produce a hundred-to-one ratio while overlooking the cash needed elsewhere in the strategy. Adding the futures reference amount to the $99 million loan would make a different mistake: treating a derivative exposure as another cash borrowing. Each position needs to be read alongside its actual cash flows.

At inception, the example has $102 million in economic assets: the bonds, the margin deposit and free cash. It owes $99 million to the lender. The difference is $3 million. A futures hedge entered at the market price contributes no initial sale proceeds. Checking those amounts prevents the diagram from creating capital that the fund never received.

The 1% haircut is an illustrative assumption. In an OFR collection from broker-dealers covering non-centrally cleared bilateral repo, zero-haircut transactions accounted for 65% of outstanding repo with hedge fund counterparties on an average daily basis, from January 2 to May 30, 2025. Scope matters. Offsetting positions or portfolio limits can provide other safeguards. A zero in one contractual field does not describe every constraint on the borrower’s leverage. ^[8]

Funding can absorb the return

Now hold the same portfolio for 90 days. Assume the combined bond, coupon and futures position generates $1.16 million before funding and fees over that period. This is an assumed combined trading result chosen to explain the arithmetic. It is neither an October 2026 market quote nor a promised return on an identified contract.

Borrowing $99 million at an average annualised repo rate of 4%, using a 360-day convention, produces $990,000 of interest over the 90 days. Add an assumed $20,000 in execution and clearing charges. The remainder is $150,000, or 5% of the $3 million committed for 90 days. We do not annualise that return. It is before the manager’s compensation, taxes and the opportunity cost of capital.

Most of the $1.16 million has gone to pay for borrowed resources. The fund retains a much smaller difference, made attractive by the limited own capital relative to the bond position. In this construction, the boundary between profit and loss depends heavily on the cost of carrying the trade.

At an average funding rate of 4.6%, the residual falls to $1,500. The break-even rate is about 4.61%, roughly 61 basis points, or 0.61 percentage point, above the starting assumption. That threshold belongs to this scenario alone, with its pre-funding result held at $1.16 million. Different relative prices, fees or coupon receipts would change it.

The maturity of the financing therefore becomes a management choice. Funding agreed through the intended exit date fixes more of the trade’s economics. Borrowings renewed along the way expose the fund to the terms available at each renewal. Practices vary across markets. The FSB’s February 2026 report describes that diversity and the constraints faced by borrowers seeking longer funding. ^[9]

For a government issuer, the implication is that a substantial source of demand can depend on a relatively narrow funding spread. Our 61-basis-point threshold cannot be applied across the sector. Funds have different contracts, hedges, liquidity resources and starting returns.

Margin calls create an immediate cash requirement

The calculation above assumes the fund gets through the 90 days. A separate scenario, independent of that period’s final return, shows what can happen along the way. Start again with the initial portfolio and its $1 million of free cash.

The bonds gain $1 million in market value. The short futures position simultaneously loses $980,000. Combined market profit is still $20,000. The hedge has offset most of the price movement. But the futures loss has to be settled in cash through variation margin. CME’s documentation describes daily cash settlement variation; the unrealised gain on the bond is not automatically deposited in the broker account. ^[7]

Add a hypothetical $500,000 increase in required initial margin. The immediate cash requirement becomes $1.48 million: a $980,000 realised futures loss and a $500,000 additional deposit. The deposit remains an asset tied up as collateral, distinct from a permanent expense. With $1 million available, the fund must find another $480,000, even though combined market profit is positive.

There are possible solutions. If the lender immediately agrees to refinance the bonds at their new $101 million value and the same 1% haircut, borrowing capacity could rise to $99.99 million. The additional $990,000 would cover the gap. An available credit line, a cash transfer or eligible cross-margining arrangements could also change the outcome. Our stress window assumes repo principal remains at $99 million: that timing assumption creates the cash shortfall.

The vulnerability concerns turning an asset into spendable money in the right account before the payment deadline. Who authorises the extra credit? Where does the transfer arrive? Which collateral has already been committed? The FSB’s work on margin-call preparedness emphasises these operational questions alongside asset quality. ^[16]

A hedge can reduce an economic loss while leaving a substantial cash requirement. Assessing resilience therefore requires both the estimated final loss and the largest payment the fund may have to make before closing the position. In our example, equity remains positive. The immediate problem is settlement.

The strategies behind the totals

How large are these positions? The OFR estimates hedge funds’ physical Treasury holdings at about $2 trillion at the end of 2025, roughly 7% of $28.9 trillion in marketable Treasuries valued at market prices. Its article was published on August 19, 2026. This estimate of long cash holdings is reconstructed from Form PF filings and CFTC futures and options data. It does not measure basis trades alone. ^[3]

Counting sellers of futures provides only part of the answer. CFTC reports break down long and short positions by participant category. A contract’s purpose depends on the rest of the portfolio. A short can hedge a purchase, form part of a relative-value trade or express an outright market view. ^[21][10]

It resembles reading a bank statement with only the debits visible. An outgoing payment tells us something happened. It does not tell us whether the money bought an asset, repaid a loan or settled a loss. A derivative observed in isolation creates the same problem: the economic context is missing.

In a note published on June 22, 2026, Phillip Monin matches physical holdings, derivatives and repo funding reported by large funds to the SEC. The resulting strategy estimates are built through sequential allocation of exposures. The author stresses that the reporting does not label individual trades. These are approximations consistent with the available information. ^[11]

CFTC research using a different form, CPO-PQR, offers another view. Better separation of some positions comes with a different reporting perimeter. Comparing studies can test an interpretation; putting their totals together as though they came from one census would manufacture a misleading series. ^[20]

Our investigation uses public tables and their stated definitions. We have neither confidential filings nor individual trading books. The estimates illuminate how the market is organised. They do not identify a named fund’s position, loss or decision to sell.

April 2025: two trades, different paths

The most revealing comparison is in the monthly data published with the Fed note. Between the end of March and the end of April 2025, the estimated cash-futures basis position rises from $725 billion to $829 billion. Estimated swap-spread positions fall from $295 billion to $237 billion. Calculated from the rounded published values, those changes are plus $104 billion and minus $58 billion, respectively. ^[12]

The swap-spread trade uses a different hedge. A fund can buy repo-financed Treasuries and enter an interest-rate swap in which it pays a fixed rate and receives a floating one. That contract changes its rate sensitivity and creates exposure to the relationship between two markets. The commitments arise under separate agreements; they lack the same bond-delivery mechanism found in the cash-futures pair. ^[13]

By the end of May, estimated swap-spread positions are down to $192 billion, a $103 billion decline from March, or about 35%. The cash-futures estimate is then $798 billion, still above its March level. Both trades involve buying bonds with borrowed money. Their published paths through this episode differ markedly. ^[12]

The statistical limits matter. A change between two estimated stocks is not a measure of total sales. Positions closed and rebuilt within a month can disappear from the final snapshot. Valuation changes, new participants and the construction of the proxies also affect interpretation. The chart challenges an undifferentiated account of all leveraged positions. It cannot establish that one strategy alone caused market stress.

This distinction changes the investigation. Futures shorts, repo borrowing and a move in bond yields can occur together without describing a general liquidation of the same trade. Identifying the hedge tells us which relative price to examine and which conditions allow a fund to continue holding it.

Selling before the lender says stop

Research on March 2020 supplies another useful corrective. Kruttli, Monin, Petrasek and Watugala find reduced arbitrage positions and cash accumulation despite relatively stable repo credit and limited contemporaneous investor withdrawals. The finding appears in their 2021 Fed working paper and in the abstract of the final article published in the Journal of Financial Economics in 2025. ^[14][15]

Average stability did not rule out pressure on particular strategies: the 2021 paper finds greater margin pressure and tighter financing terms for funds predominantly trading the basis. ^[14]

A manager can therefore reduce a portfolio before a bank withdraws financing. The logic is familiar. Anticipating that volatility will require more cash, the manager may prefer to release it now. That decision protects the fund while reducing the market’s capacity to carry bonds. Similar decisions can arrive together without any common instruction.

The distinction matters for policy. More available bank credit may address a refused loan. It does not ensure that a fund will keep a position once its own risk limits or expected liquidity needs make it unattractive. The availability of a resource and the willingness to use it are separate decisions.

The 2020 evidence has a limit of its own. Observed conditions developed during a crisis accompanied by public intervention. They do not show what repo credit would have looked like without that support. The research challenges an automatic explanation based on a general funding shutdown in the sample studied; it does not promise stable financing during the next shock. ^[14]

Concentration also matters. In Monin’s sample, the largest 50 funds account for about 90% of gross Treasury exposures in September 2025. The denominator is the funds’ exposure, including different position types. It is neither their share of US government debt nor a measure of potential losses. ^[11]

A conditional creditor

Europe is part of this structure too. In its May 2026 Financial Stability Review, the ECB distinguishes the comparatively small hedge fund sector domiciled in the euro area from global funds’ significant activity in its sovereign markets. Legal domicile and the location of risk describe different geographies. The US estimates used here cannot quantify those funds’ exact share of French OATs. ^[19]

A more recent observation comes from the ECB survey published on October 7, 2026, with responses from 26 large banks. For June to August, respondents reported slightly higher hedge fund leverage and slightly tighter non-price credit terms for these funds, despite an overall easing in credit conditions. These qualitative assessments cover euro-denominated securities financing and OTC derivatives; they measure neither OAT holdings nor individual funds’ sales. ^[23]

The industry’s strongest argument should be kept in view. Connecting markets and carrying securities supplies a useful intermediation service. Constraining that capacity can shift activity or make the service more expensive. This is central to MFA’s response to the IMF. The precise effect on government funding costs must still be established for each proposed measure. ^[17]

The Bank of England’s April 2026 consultation feedback also notes limited quantitative evidence behind some objections concerning costs or activity moving elsewhere. That calls for documented comparisons rather than an assumed mechanical effect from every minimum haircut. Additional capital has a cost. So do the forced sales it may help prevent. ^[18]

On July 17, 2026, Sarah Breeden, the Bank of England’s Deputy Governor for Financial Stability, explained that minimum haircuts could take portfolio-wide risk into account. In her assessment, a well-calibrated approach need not raise total margin costs where overall collateral is already adequate. The proposal remains under discussion; its effects would depend on the eventual design. ^[22]

The central issue is the duration of the investor’s commitment. A fund can hold a long-dated bond while intending to carry it only as long as its combination of financing, hedging and liquidity remains acceptable. The Treasury’s credit quality leaves that timing question unresolved.

Measuring the reliability of this demand requires looking beyond the amount bought. The hedge describes part of the price risk. Funding contracts establish maturities and rights. Resources that can actually be transferred determine whether the fund can wait. Public reporting reveals pieces of the structure, with delays and gaps that remain important.

Part four follows the money that makes waiting possible: the funds’ lenders, the collateral they receive and the obligations they retain in turn. The bond has reached a creditor. Its financing continues the journey.

Method and scope

Documentary investigation closed on October 11, 2026. Public series are dated by their statistical periods, separately from publication dates. The dollar amounts, rates, fees and margin calls in the examples are l0g assumptions. The 90-day calculation and intraday scenario are separate. No interviews, private portfolios or market returns have been invented. The references below link to the public data and methods; assumptions and calculations are set out in the article. Accessible working-paper versions and the publisher’s abstract are identified separately.

Sources and reading notes

1. [1] US Treasury / Treasury Borrowing Advisory Committee. Discussion of Treasury Futures Positions Across Different Investor TypesPublication : January 30, 2024. Period / scope : Presentation dated 30 January 2024; structural discussion, not a current portfolio census.https://home.treasury.gov/system/files/221/TBACCharge1Q12024.pdf Futures demand can reflect duration management, allocation choices and operational constraints. Slides 10 and 17 discuss these uses.
2. [2] International Monetary Fund. Hedge Funds Improve Market Functioning, but Can Also Amplify StressPublication : October 6, 2026. Period / scope : Publication on 6 October 2026; combines different asset classes and historical samples.https://www.imf.org/en/blogs/articles/2026/10/06/hedge-funds-improve-market-functioning-but-can-also-amplify-stress The blog discusses market-functioning benefits and the interaction of leverage, crowded positions and redemptions. Only the blog was used, not an assumed reading of the full GFSR chapter.
3. [3] Office of Financial Research; Ted Berg and Daniel Stemp. Hedge Funds’ Cash Treasury Holdings Reach $2 TrillionPublication : August 19, 2026. Period / scope : End-2025 estimated cash Treasury holdings; publication 19 August 2026.https://www.financialresearch.gov/the-ofr-blog/2026/08/19/hedge-funds-cash-treasury-holdings/ Estimated physical holdings are distinguished from derivative exposures. The $28.9tn denominator is the market value of outstanding marketable Treasuries.
4. [4] CME Group. The Basics of Treasuries BasisPublication : date not established. Period / scope : Technical definitions; page consulted on 11 October 2026.https://www.cmegroup.com/education/courses/introduction-to-treasuries/the-basics-of-treasuries-basis Gross basis equals clean cash price less futures price times its conversion factor. This convention differs from a net arbitrage return.
5. [5] CME Group. Calculating U.S. Treasury Futures Conversion FactorsPublication : January 19, 2024. Period / scope : Technical contract explanation, published January 2024.https://www.cmegroup.com/articles/2024/calculating-us-treasury-futures-conversion-factors.html Different coupons and maturities can be deliverable. A contractual factor adjusts the delivery price.
6. [6] CME Group. Margin: Know What’s NeededPublication : date not established. Period / scope : Technical definitions; consulted on 11 October 2026.https://www.cmegroup.com/education/courses/introduction-to-futures/margin-know-what-is-needed Initial margin protects against default risk and can change. Brokers may impose additional requirements.
7. [7] CME Group. Money Calculations for CME-cleared Futures and OptionsPublication : June 11, 2015. Period / scope : Technical document updated 11 June 2015.https://www.cmegroup.com/clearing/files/CME-Money-Calculations-Futures-and-Options.pdf Futures settlement variation is paid in cash each day and differs from the initial-margin deposit.
8. [8] Office of Financial Research; Ashlyn Cenicola, Robert Mann and Mark Paddrik. Are Zero-Haircut Repos as Common as Advertised?Publication : August 12, 2025. Period / scope : Daily non-centrally cleared bilateral repo data, 2 January to 30 May 2025.https://www.financialresearch.gov/the-ofr-blog/2025/08/12/are-zero-haircut-repos-as-common-as-advertised/ Zero haircuts account for 65% of the hedge-fund-counterparty repo segment studied. Portfolio offsets and other limits can still apply.
9. [9] Financial Stability Board. Vulnerabilities in Government Bond-backed Repo MarketsPublication : February 4, 2026. Period / scope : Global review drawing substantially on 2024 market data and information from member jurisdictions.https://www.fsb.org/uploads/P040226.pdf Repo entails contractual sale and repurchase of the security. Funding horizons differ across lenders, markets and agreements.
10. [10] Federal Reserve; Jonathan Glicoes, Benjamin Iorio, Phillip Monin and Lubomir Petrasek. Quantifying Treasury Cash-Futures Basis TradesPublication : March 8, 2024. Period / scope : Methodological note with historical positions and measures through its publication.https://www.federalreserve.gov/econres/notes/feds-notes/quantifying-treasury-cash-futures-basis-trades-20240308.html Profitability combines purchase, delivery or close-out, coupons, funding and delivery options. A futures short total does not identify the strategy by itself.
11. [11] Federal Reserve; Phillip J. Monin. Decomposing Hedge Funds’ U.S. Treasury ExposuresPublication : June 22, 2026. Period / scope : Monthly large-hedge-fund observations through September 2025.https://www.federalreserve.gov/econres/notes/feds-notes/decomposing-hedge-funds-u-s-treasury-exposures-20260622.html Proxies combine physical bonds, derivatives and funding. The top 50 funds account for about 90% of gross Treasury exposures in the sample, not 90% of government debt.
12. [12] Federal Reserve; accessible data accompanying Phillip J. Monin. Decomposing Hedge Funds’ U.S. Treasury Exposures, Accessible DataPublication : June 22, 2026. Period / scope : Excerpt from figure 3, January 2023–September 2025, USD billions. The full table begins in January 2013.https://www.federalreserve.gov/econres/notes/feds-notes/decomposing-hedge-funds-u-s-treasury-exposures-accessible-20260622.htm March–April 2025: basis 725 → 829; swap spread 295 → 237, in $bn. March–May: swap spread 295 → 192. Changes use the published rounded integers.
13. [13] Bank for International Settlements; Vladyslav Sushko and Karamfil Todorov. Sizing up hedge funds’ relative value trades in US Treasuries and interest rate swapsPublication : December 2025. Period / scope : December 2025 Quarterly Review; historical evidence through 2025.https://www.bis.org/publications/qr-202512/sizing-hedge-funds-relative-value-trades-us-treasuries-and-interest-rate-swaps The position combines a bond with a pay-fixed, receive-floating swap. Its exit mechanics differ from delivery of a bond into a future.
14. [14] Federal Reserve; Mathias S. Kruttli, Phillip J. Monin, Lubomir Petrasek and Sumudu W. Watugala. Hedge Fund Treasury Trading and Funding Fragility: Evidence from the COVID-19 CrisisPublication : June 2021. Period / scope : March 2020 crisis; FEDS 2021-038, not 2026 fund behaviour.https://www.federalreserve.gov/econres/feds/hedge-fund-treasury-trading-and-funding-fragility-evidence-from-the-covid-19-crisis.htm Position reductions coexist with relatively stable repo funding and limited contemporaneous redemptions. Precautionary liquidity and internal risk constraints matter.
15. [15] Journal of Financial Economics; Kruttli, Monin, Petrasek and Watugala. LTCM Redux? Hedge fund Treasury trading, funding fragility, and risk constraintsPublication : July 2025. Period / scope : Journal of Financial Economics 169, article 104017; study of March 2020.https://www.sciencedirect.com/science/article/pii/S0304405X2500025X The final abstract confirms cash accumulation and reduced arbitrage despite maintained credit and limited withdrawals.
16. [16] Financial Stability Board. Liquidity Preparedness for Margin and Collateral Calls: Final reportPublication : December 10, 2024. Period / scope : Policy recommendations published 10 December 2024; past stress cases and preparedness.https://www.fsb.org/uploads/P101224-1.pdf The report emphasises resources that can actually be mobilised, operational processes and liquidity stress scenarios, including for high-quality assets.
17. [17] Managed Funds Association; Jillien Flores. MFA statement on IMF Report on hedge funds and financial stabilityPublication : October 6, 2026. Period / scope : Industry statement responding to the IMF on 6 October 2026.https://www.mfaalts.org/statement/mfa-statement-on-imf-report-on-hedge-funds-and-financial-stability/ MFA defends fund participation and challenges blanket restrictions. This is presented explicitly as the industry’s position.
18. [18] Bank of England. Enhancing the resilience of the gilt repo market: discussion paper feedback statementPublication : April 1, 2026. Period / scope : Feedback published April 2026 on the September 2025 discussion paper.https://www.bankofengland.co.uk/paper/2026/discussion-paper/enhancing-the-resilience-of-the-gilt-repo-market-discussion-paper-feedback-statement Consultation feedback records concerns about costs and activity moving elsewhere, with limited quantitative evidence to measure them.
19. [19] European Central Bank. Financial Stability Review, May 2026Publication : May 27, 2026. Period / scope : Review published May 2026; different underlying cut-off dates by section.https://www.ecb.europa.eu/press/financial-stability-publications/fsr/html/ecb.fsr202605~50566915a7.en.html The ECB distinguishes the relatively small euro-area-domiciled sector from global funds’ activity in European sovereign debt.
20. [20] CFTC Office of the Chief Economist; Scott Mixon and Alexei Orlov. Observations on the Treasury Cash-Futures Basis TradePublication : September 23, 2024. Period / scope : OCE Working Paper 2024-007; historical sample of selected funds.https://www.cftc.gov/sites/default/files/Basis_trade_Mixon_Orlov_ada.pdf CPO-PQR reporting provides a different view of cash bonds and futures. The studies have different samples and perimeters.
21. [21] Commodity Futures Trading Commission. Commitments of Traders: Traders in Financial FuturesPublication : date not established. Period / scope : Methodology page consulted 11 October 2026; no current position totals used.https://www.cftc.gov/MarketReports/CommitmentsofTraders/index.htm An investor category and the direction of a position do not describe the other legs of the portfolio.
22. [22] Bank of England; Sarah Breeden. Gilt edged resilience: strengthening liquidity provision in the repo marketPublished: July 17, 2026. Scope: proposed gilt repo reforms, still under discussion.https://www.bankofengland.co.uk/bank-insights/2026/gilt-edged-resilience-strengthening-liquidity-provision-in-the-repo-market Breeden argues for an approach that considers the full portfolio and existing collateral. This is a policy argument, not an enacted requirement.
23. [23] European Central Bank. Results of the September 2026 survey on credit terms and conditions in euro-denominated securities financing and OTC derivatives markets (SESFOD)Published: October 7, 2026. Scope: June–August 2026, qualitative responses from 26 large banks.https://www.ecb.europa.eu/press/pr/date/2026/html/ecb.pr261007~6447350434.en.html Slightly higher hedge fund leverage and tighter non-price terms alongside an overall easing. This does not measure holdings of French debt.