British banks are delivering more credit-card, vehicle-finance and other less-liquid assets to the Bank of England in return for reserves. That sounds like a warning until it is placed inside a deliberate redesign of Britain's monetary plumbing. The Bank is shrinking reserves previously created through asset purchases and expects regular repo facilities to supply more of the stock instead.
The Reuters investigation nevertheless identifies a real risk question. An operating framework designed for routine use can still move harder-to-value collateral onto the central bank's balance sheet. The amount alone cannot diagnose stress; its auction price, concentration and performance determine how much the signal should concern investors.
One auction is striking, the stock is more informative
Banks received £1.9 billion against Level C collateral in the 18 August 2026 Indexed Long-Term Repo auction, according to Reuters' review of Bank filings. That was the largest weekly amount since March 2020 and three times the prior week's total. Reuters calculated that roughly £17.8 billion of Level C collateral was outstanding through the ILTR, up from £8.7 billion a year earlier and less than £1 billion in mid-2024.
The longer baseline softens, but does not erase, the change. The Bank's official market-operations report recorded £69.9 billion of total ILTR drawings at the end of February 2026: £50.6 billion against Level A, £5.3 billion against Level B and £14.1 billion against Level C. Seventy-nine firms had drawings, compared with 43 a year earlier.
Reuters' £17.8 billion estimate is therefore about £3.7 billion above that February Level C stock, rather than a fresh leap from £8.7 billion. Both comparisons are factual, but they answer different questions. The year-on-year figure captures the framework's rapid expansion; the February comparison better isolates the recent increment.
Repo supply is replacing a shrinking reserve source
The increase is partly the product, not a malfunction, of policy. As government bonds bought under quantitative easing mature or are sold, the reserves created to pay for them decline. The Bank wants demand-driven repos to provide the quantity banks need for payments and monetary control instead of maintaining an oversized bond portfolio.
The Prudential Regulation Authority made the intended behaviour explicit in its ILTR statement. It said the facility should be used freely, that participation should be viewed as routine sterling liquidity management, and that firms should practise using it. The Bank later reported that total ILTR use had broadened to roughly £70 billion across about 80 firms.
This policy is designed to remove stigma. If every drawing were interpreted as distress, healthy banks would avoid the facility until they had no alternative, undermining its role as routine infrastructure. Rising participation and stock are therefore expected as the system changes.
Level C buys a liquidity upgrade at a price
Level C is not a synonym for a bad loan. The Bank's Market Operations Guide says it typically includes less-liquid securitisations, assets delivered by the entity that originated the underlying exposure and portfolios of loans such as mortgages. Firms keep the credit exposure economically relevant, but temporarily exchange eligible collateral for central-bank reserves.
The ILTR normally runs weekly, provides funds for six months and settles two business days after the auction. Its supply curve allocates reserves by the spread bidders offer over Bank Rate. The participant guide says Level C drawings would typically be supplied around 20 to 40 basis points above Bank Rate when system reserve demand is in the expected range, more expensively than liquid Level A collateral. Total auction capacity can expand to £35 billion if bids justify it.
Pricing and haircuts are the control mechanism. A bank pays for the liquidity upgrade, while the central bank discounts collateral values and can demand enough assets to protect itself if a counterparty fails. Greater Level C use raises valuation, legal and liquidation work, but eligibility does not mean the Bank assumes the loans at face value.
Routine use does not make the signal empty
The strongest reassuring interpretation is structural: the Bank told firms to pre-position broad collateral and use the facility, then use rose. The £14.1 billion February baseline and wide participation support that reading.
The counterargument is that routine infrastructure can also carry information. A firm may choose Level C because private funding against the same assets is costly, because it wants to preserve liquid securities, or because its funding position is tightening. Published results are aggregate; the Bank does not identify counterparties or disclose individual transactions. A benign sector total can therefore conceal concentrated demand.
Nor does an accepted haircut eliminate all risk. Models can misestimate default correlation, recovery timing or market depth during stress. The concern is not that every underlying store-card or vehicle loan will fail, but that assets which look diversified in normal conditions may become harder to finance or sell together.
Spreads and concentration would settle the argument
Future auction data should be read as a bundle. Stable clearing spreads, broad bidder participation and Level C growth proportional to falling system reserves would support the routine-transition explanation. Abruptly wider spreads, repeated large bids, a shrinking bidder count or changes to collateral haircuts would make funding pressure or risk reassessment more plausible.
Better disclosure of collateral concentration and asset performance would also help, even without naming firms. Until that evidence changes, the balanced conclusion is that more Level C usage is neither a crisis siren nor an empty statistic. It shows the repo-led framework reaching assets beyond sovereign bonds, exactly as designed, while moving the quality of the Bank's risk controls closer to the centre of monetary implementation.