The UK refinancing wall, 2026–2028.

A queue of maturing debt, papered between 2020 and 2024.

The refinancing wall — often called the maturity wall — is the concentration of company debt falling due for renewal across 2026 to 2028. Facilities written in the pandemic-era lending of 2020, the recovery lending of 2021 and the repricing wave of 2022 to 2024 were mostly signed on customary three-to-five-year senior terms, so they age into their refinancing windows together. Facilities papered in one period mature in one period: that is the whole mechanism, and it is why the wall does not require anything further to happen. It only requires time to pass.

Most published estimates of the wall are built from the top down, from scheme-lending vintages, leveraged-loan samples or survey sentiment. Ours is built from the bottom up. Companies House publishes every security interest a UK company grants (who holds it, when it was created, what it covers), and we scored 83,957 lower-mid-market companies against that register. 26,383 of them carry an outstanding registered charge whose age implies an estimated refinancing window; 14,675 were inside that window at 7 July 2026. The full analysis, with charts, is in the flagship research piece, and the method is summarised at the foot of this page.

The wall is a plateau, not a spike.

The estimated windows are years wide, so the honest measure is how much paper sits inside its refinancing zone at any moment rather than how many facilities mature in a given quarter. On that measure the wall reads as a plateau: between 15,519 and 16,808 estimated windows are open in every quarter from the start of 2026 through mid-2027, peaking at 16,808 in the second quarter of 2027, before tapering to 9,282 by the end of 2028. A spike is a scheduling problem; a plateau is a capacity problem, because the same lender credit teams, advisers and lawyers have to clear an elevated volume for eight consecutive quarters. A borrower who assumes the crowd thins out next year is misreading the shape: the pressure holds through mid-2027.

Five banking groups hold more than half of it.

Among the 14,675 in-window companies, 13,513 carry a named incumbent (1,642 distinct charge-holding entities), but the distribution is anything but flat. Folded to banking groups, five names hold 52.5% of the primary charges, and clearing banks as a category hold 54.9%. A further 1,162 charges sit with bare security trustees, where a real facility is implied but the lender behind it is not disclosed, and 3,831 with corporate holders the register does not name in a recognisable form; both are counted in every denominator, so the named shares are floors.

Both of the obvious readings of that concentration are wrong. It does not show that banks are the wrong home for this paper; for most of these borrowers a bank facility was, and often still is, the cheapest and simplest answer, and plenty of these refinancings should end exactly where they started. Nor is the concentration harmless: when five credit desks face half the wall at once, each is triaging, and its pricing on any given renewal reflects its book, its sector appetite that quarter and its capacity — none of which the borrower can see. The consequence is the same whichever way an individual decision goes: an in-window borrower has more to gain from a genuine competitive process than in a normal year, even when the best outcome is the incumbent matching a better term sheet.

My loan matures in 2026. What should I do?

Start now, and start with the dates. A 2026 maturity sits in the densest stretch of the plateau: 16,598 estimated windows are open in the final quarter of 2026 alone, and the crowd does not thin until after mid-2027, so there is no quieter market to wait for. Your facility agreement, not the register, is the authority on your own dates. Pull it and confirm the contractual maturity, any extension options and the notice periods they carry; then check what the market can see about you, which is the charge on your Companies House file and its creation date.

Then build the pack the process will demand before anyone asks for it: three years of filed accounts, current management information, a forecast that shows the debt being serviced through a downside case, and the covenant calculations that go with it. That pack is the price of admission with every lender category, and having it ready is what lets you run the renewal conversation and the wider market approach in parallel rather than in sequence. How much debt a lender will hold against your numbers is its own question, answered in the borrowing-capacity guide.

The sequence is the part borrowers get wrong. Going to the incumbent first, waiting for its answer and only then looking wider hands the clock to the other side of the table. Open the market while the renewal conversation runs and let each inform the other; on a wall where every desk is triaging, the file that arrives complete, with a live alternative behind it, is the one that gets priced with care.

What if my bank will not renew?

A bank that does not want a renewal rarely refuses it outright. The signal arrives sideways: a short extension offered where a new term was expected, a margin set high enough to lose the business politely, a request for security or guarantees the facility never carried, or a relationship manager who begins mentioning the merits of exploring options. Read those as a decision, because that is what they are. On this wall the decision is often about the desk rather than the borrower: five groups face half the in-window cohort, and a book that is full of your sector this quarter can pass on a credit it would have renewed without a second look in a quieter year.

Treat the signal as a timetable, not a verdict on the credit. The register counts 1,667 distinct charge-holding entities across the in-window cohort alone, and they sort into categories with different answers to the same borrower: other clearing and specialist banks, the keenest pricing where the leverage and the story fit bank appetite; asset-based lenders, where the balance sheet carries receivables, stock or plant to lend against directly; and debt funds, which will hold more leverage and stay with a harder story at a materially higher price, a trade to be priced deliberately rather than accepted as the only open door. The comparison guide sets that trade out in full.

There is a harder reading to hold alongside the hopeful one. If the incumbent steps back and every other category quotes wide, the market is commenting on the amount rather than the address, and the honest response may be a smaller facility, a structure built on the assets, or equity rather than debt for the piece no lender will hold. That conclusion is rarer than borrowers fear, but a process that cannot reach it is not a process; it is shopping for validation. The time to find out which reading applies is twelve months before maturity, not six.

Aggregate register data, stated with its limits.

The method in one paragraph: each outstanding registered charge is assigned a customary facility tenor from its type and holder (a clearing-bank or challenger debenture reads as a four-year facility give or take eighteen months, a debt-fund charge as a six-year unitranche) and the estimated refinancing window is the creation date plus that tenor. Rolling facilities, invoice finance and ABL are excluded because they carry no refinancing clock; rent-deposit, pension and landlord charges are excluded because they are not facilities. One company, one primary window. Every window is an estimate, typically three years wide; a charge proves security was granted, not that the facility is still outstanding; and the 83,957 companies are a prioritised sample of a much larger universe, so every count is a floor. The cohort sizes, the origination vintages and the concentration of holders are the hard part of the data, and they are what this page leans on.

The four-year assumption is the obvious thing to attack, so we measured it. A satisfaction filed within ninety days of a new charge at the same company is a refinancing: the incoming lender requires the old security released before it will draw, so the paperwork is prompt and the dates are reliable. Across 55,482 of those, the median life of a facility is 3.9 years, against the four years the model assumes. The quartiles are 1.9 and 6.7 years, which is wider than the eighteen-month half-width used here, so if anything these windows are drawn slightly narrow.

The same measurement says something worth knowing whatever your maturity is. Satisfactions with no refinancing behind them run 1.2 years longer, and across every charge on the file the median register life is 11.8 years. Charges are released late or not at all, so an outstanding charge is weak evidence that the debt behind it is live. It will still be found in diligence, and it will still have to be cleared before a new lender draws. Reading your own charge register covers what to do about it.