The retention book is the most profitable money a contractor never has. On a healthy job you have already earned it — the work is done, the margin is in it — but roughly 5% of each valuation is withheld, half released only at practical completion and the rest after the defects liability period, which runs twelve months on most jobs and up to twenty-four on residential schemes. Across a busy order book that is a permanently parked balance, growing as turnover grows. Funding it badly is how profitable construction businesses run out of cash.
Why the gap is structural, not a one-off squeeze
Construction pays in arrears against staged certificates, so a contractor banks each interim valuation already net of retention and already net of any CIS deduction the payer makes at source. The cash that arrives is always less than the value certified, and the difference does not catch up until completion and final account. Run several projects at once and the withheld slices stack: you can be trading profitably on paper while the entire retention book — real, earned, undisputed money — sits eighteen months out of reach.
That is a different problem from a slow debtor. A late invoice gets chased and paid. Retention is contractually locked for a fixed window, so it cannot be accelerated by better credit control. It can only be funded, released early by agreement, or protected so that it actually comes back.
The reform pressure is real, but do not build cashflow on it
The policy direction is finally moving, which matters for how you fund the next two years rather than the last ten. From 1 March 2025 larger businesses — broadly, those over £36m turnover or 250 staff — must disclose their retention practices under the amended payment-reporting regulations: whether they withhold, at what rate, and on what terms, making the sector's reliance on withheld supply-chain cash visible for the first time. A private UK Retention Deposit Scheme launched in January 2025 to hold withheld sums in a ring-fenced, protected form rather than on the payer's balance sheet. And in July 2025 the government opened a consultation on reforming or banning cash retentions outright, running into the autumn.
The honest read: protection and transparency are arriving, an outright ban is not yet law, and none of it releases the retention sitting in your ledger today. Treat reform as a reason to insist on ring-fenced retention on new contracts — not as a funding plan for the cash already trapped.
Match the funding to where the cash is actually stuck
The instinct under a cash squeeze is to reach for one more facility and gear the whole business against it. That is the trap. The retention gap and the plant burden are two different problems and they want two different structures.
Where the trapped capital is plant — excavators, telehandlers, site accommodation, the yellow kit that ate the cash in the first place — the answer is rarely a working-capital loan. Asset finance, and specifically sale-and-leaseback of kit you already own outright, releases equity against a depreciating asset at a rate set by the asset, then spreads the cost across the jobs the kit will earn on. You match a long asset to long funding instead of borrowing short and unsecured to cover a hole that plant created.
Where the gap is genuinely the retention and the certified-payment lag — work done, cash earned, calendar not yet caught up — the right tool is a cashflow facility sized to the gap, not the turnover. A cashflow loan bridges the certified-payment cycle and known retention releases without taking a charge over every asset on the yard, which keeps the plant free to refinance separately when the next contract start needs it.
For a contractor whose value is spread across debtors, the retention book, stock and plant at the same time, wrapping the lot into a single revolving line through asset-based lending usually lends deeper than three single-product facilities stacked on top of each other — and a specialist ABL desk will treat the retention element as its own line rather than pretending it is ordinary debtors.
What over-gearing looks like before it bites
The failure mode is predictable. A contractor covers a retention-driven gap with short, expensive money — a merchant-cash-style advance or a personally guaranteed unsecured loan — then takes the next contract, which carries its own retention, which deepens the gap the loan was meant to close. Each new job makes the cash position worse, not better, because growth multiplies the withheld slice. By the time the first retention releases, the facility that bridged it has rolled twice and cost more than the retention was worth.
Avoiding that turns on two questions, asked before the facility not after. First: is this gap the plant or the retention — because the structures are not interchangeable, and using a cashflow line to fund kit, or asset finance to fund a payment lag, leaves you geared against the wrong thing. Second: what is the realistic retention recovery — some of it never comes back, and a facility sized against gross retention is over-lent the day it draws. Bedrock prices the gap against retention the way a careful underwriter does, discounted to what actually returns, and points contractors at the construction sector page to map which of those structures fits the cash that is genuinely stuck. The cheapest facility is the one matched to the asset behind it. The most expensive is the one taken because it was the only call you made.