How much can your business actually borrow?
There is no single number, but there is a single method. A lender sizes your debt capacity through three lenses: a multiple of your earnings, what your cash flow can service, and what your assets would support — and the answer is the lowest of the tests that apply to the structure you choose. Most owners start from the multiple, because it is the number the market talks about: a bank at two-and-a-half to three-and-a-half times EBITDA, a fund stretching to four or more. But at 2026 interest rates the multiple is often not the test that binds. On many deals the affordability arithmetic — interest and repayments against the cash the business actually generates — caps the number first, and for an asset-rich business the balance sheet can carry more than the earnings ever would. This guide works through all three lenses, at July 2026 market rates, and then through the harder question: how much of what you can raise you should actually carry.
Written for the borrower’s side of the table. This is general guidance, not advice on your specific facility. It deepens the shorter answers in our working guide.
A multiple of EBITDA — but of the EBITDA that survives diligence.
Cash-flow lending is sized as a multiple of EBITDA, and the multiple moves with the lender type. On senior cash-flow terms a bank will typically lend around 2.5 to 3.5 times EBITDA to a decent lower-mid-market business — the top of that range for a clean, cash-generative credit, the bottom for anything with cyclicality or concentration in it. A unitranche fund will usually stretch to 4 to 4.5 times, occasionally to five for a strong, sponsor-backed credit with recurring revenue. The stretch is not generosity; it is a different risk appetite at a materially higher coupon, and whether the extra turn is worth its price is the subject of our guide to unitranche versus bank senior debt.
The number the multiple is applied to matters as much as the multiple itself, and it is not the EBITDA in your management accounts. Every lender scrubs the figure. Add-backs are tested one by one: genuinely one-off costs survive; “exceptional” items that recur every year do not, and nor do owner adjustments a diligence accountant cannot evidence. Where the raise is large enough to justify it, a quality-of-earnings review does this formally, testing how the earnings were built, how sustainable they are and how much of the reported profit turns into cash. The arithmetic is unforgiving: at three times, £300,000 of disallowed add-backs is £900,000 of borrowing capacity gone. A borrower who walks in with a £2m EBITDA and walks out of diligence with £1.7m has not lost £300,000 — they have lost the better part of a million pounds of debt.
The practical discipline follows directly: scrub your own EBITDA before a lender does. Build the bridge from statutory profit to adjusted EBITDA yourself, evidence every add-back, and drop the ones you cannot defend. A defensible number presented plainly borrows more than an optimistic one taken apart in diligence, because the lender prices what it cannot verify as risk.
At 2026 rates, affordability often binds before leverage.
The second lens is the one the first ignores: whatever the multiple says, the cash flow has to carry the payments. Mid-market debt is floating-rate, priced over SONIA, which sat at 3.73% on 2 July 2026 against a Bank Rate of 3.75%, held at the Bank of England’s June meeting. Bank of England, Bank Rate. On top of that base, UK mid-market unitranche is broadly SONIA plus 550 to 800 basis points — an all-in coupon of roughly 9.25% to 11.75% before fees. Deloitte Private Debt Deal Tracker. The arithmetic that follows is simple enough to do on the back of the term sheet: the share of EBITDA that interest consumes is the leverage multiplied by the rate. Four turns of debt at an all-in of about 10.25% means interest alone takes roughly 41% of EBITDA, every year, before tax, capex or a pound of principal. Interest cover — EBITDA over the interest bill — sits at about 2.4 times. That is servicing, but it is not comfort.
Now run it the other way, which is how a credit committee runs it. Cash-flow lenders test interest cover as a covenant, on more levered structures commonly set somewhere around two times, and a sensible borrower wants headroom of 25 to 30% against the base case rather than a plan that starts on the line. Hold cover at two-and-a-half times — a two-times test with real headroom — and the interest bill can be at most 40% of EBITDA. At a 10.25% all-in rate, that caps the debt at about 3.9 times EBITDA: below the 4 to 4.5 times the leverage lens advertised. This is the quiet consequence of the rate environment. When money cost 6%, a 2.5-times cover floor allowed nearly 6.7 turns of debt and the leverage multiple was always the binding test; at today’s rates the two lenses cross, and on many deals the affordability arithmetic caps the number before the multiple does.
On bank debt the same lens binds through a different line, because bank term loans amortise. A loan repaid over three to five years adds principal on top of interest, and the test becomes cash-flow cover: can the business fund the whole debt service, interest and repayments together, out of the cash it generates after tax and capex. That is why a bank lends 2.5 to 3.5 times rather than more — not timidity, but the arithmetic of getting repaid inside the term — and why amortisation profiles are shaped, part amortising and part final balloon, to fit the cash the plan actually throws off. The full pricing stack behind these rates, fees and margins line by line, is in our guide to the all-in cost of raising debt.
Where the balance sheet is heavier than the earnings, it can out-lend the multiple.
The third lens ignores EBITDA multiples altogether. An asset-based lender sizes a facility from a borrowing base: the sum of the assets it will advance against, each at its own rate. The largest line is almost always the debtor book, at typically 80 to 90% of the eligible receivables — the highest advance rate of any asset class, because trade receivables are short-dated, self-liquidating and readily valued. Below that sit a lower percentage tranche against stock and, where the business owns them, valuation-based advances against plant and machinery and against property. The word doing the work is eligible: overdue and disputed invoices come out, a concentration cap limits how much any single customer can represent, and slow-moving stock is written down, so the eligible base is routinely smaller than the gross one.
For the right balance sheet this lens produces a bigger answer than either of the first two. A distributor or manufacturer running thin margins on high turnover can carry a modest EBITDA on top of a large debtor book and a warehouse of stock; a multiple of the earnings sizes a small loan, while the borrowing base against the same balance sheet can release considerably more, and it flexes upward as the book grows. The reverse is equally true, and worth saying plainly: an asset-light services business has almost nothing for this lens to measure, and for it the cash-flow lenses are the only ones that apply. The mechanics — the base certificate, the field audits, the reporting discipline the availability is priced on — are set out in our guide to asset-based lending.
The three lenses are not exclusive. Blended structures are common in the lower mid-market: an asset-based line against the working-capital core with a cash-flow tranche on top, or a term loan sized off EBITDA with a receivables facility alongside it. Which combination releases the most, at the lowest all-in cost, is a question of the shape of your balance sheet against the shape of your earnings — which is why the honest answer to “how much can I borrow” starts with which lens flatters your business least.
Two businesses with the same EBITDA can borrow very different amounts.
The multiples in lens one are ranges, and where you land within them — or whether a lender will engage at all — turns on the quality of the earnings behind the number. Recurring revenue moves capacity up more than any other single factor: contracted, repeatable income lets a lender underwrite the forward cash flow with confidence, which is why the top of the unitranche range is reserved for credits with real revenue visibility, and why a business sold on one-off projects sits at the bottom of the bank’s. Customer concentration moves it down: a business where one customer is a third of revenue carries a risk no covenant can fix, and every lens penalises it — the cash-flow lender through a lower multiple, the asset-based lender through a concentration cap on the borrowing base.
Cyclicality lowers the number because the lender underwrites the trough, not the peak: a business whose EBITDA swings with the cycle is leveraged off a through-cycle view of its earnings, and debt sized at the top of a good year is the classic way a manageable facility becomes an unmanageable one. Cash conversion decides how much of the EBITDA is real for the purpose of servicing debt. A business that converts 90% of EBITDA into cash can carry more than one that leaks it into working capital and capex, whatever the multiple says, because the affordability lens runs on cash rather than on an accounting line. Sector sets the frame around everything else: lenders run appetite by sector, and the same numbers borrow more in a sector a credit committee understands and likes than in one it has been burned by. None of these is fixed in the short term, but all of them are worth knowing before you ask, because they explain the gap between the headline multiple and the offer that arrives.
One business, three answers.
Take a business with £2m of EBITDA — scrubbed, defensible, post-diligence — and run it through each lens. These figures are illustrative arithmetic at July 2026 rates, not a quote; the real number turns on your credit, your sector and the process.
Lens one — the multiple
On bank senior terms at 2.5 to 3.5 times, the multiple supports £5m to £7m. On unitranche at 4 to 4.5 times, £8m to £9m — with five times, £10m, reserved for an exceptional credit. The spread between the conservative bank case and the full fund stretch is £5m on the same EBITDA, which is why “how much can I borrow” has no answer until you say from whom, and at what price.
Lens two — affordability
Now test what the cash flow carries. At four times — £8m of unitranche — an all-in rate of about 10.25% means an interest bill of roughly £820,000 a year: 41% of EBITDA, with cover at about 2.4 times. Across the full 9.25% to 11.75% pricing range the bill runs £740,000 to £940,000 — 37% to 47% of EBITDA — before tax, capex or any principal. Hold cover at a prudent two-and-a-half times and the interest bill must stay under £800,000, which at 10.25% caps the debt at about £7.8m — call it 3.9 times, under the 4 to 4.5 times the multiple offered. On the bank alternative, £6m at three times and an all-in near 6.75% costs about £405,000 a year in interest — cover of nearly five times — but the loan amortises: repaid straight-line over five years it adds £1.2m of principal, and £1.6m of total annual debt service against £2m of EBITDA does not work once tax and capex are paid. The bank deal is affordable at that quantum only with a longer profile or a balloon, which is exactly the negotiation the term sheet exists for.
Lens three — the asset base
Suppose the same business carries £5m of trade receivables and £2m of stock. An asset-based lender advancing 80 to 90% of the eligible debtor book releases £4m to £4.5m against the receivables — assuming the whole book survives the eligibility rules, which it rarely quite does — plus a smaller tranche against the stock. For this business the asset lens lands below the cash-flow lenses, and ABL is a working-capital complement rather than the main answer. Give the same £2m of EBITDA a £12m debtor book instead, and the ranking reverses: the base out-lends the multiple, and the asset route becomes the one to price first.
The summary the three lenses produce: this business can plausibly raise £5m to £7m from a bank at a price the cash flow carries comfortably, or push toward £8m with a fund at a price that consumes two-fifths of its EBITDA in interest. The capacity question has been answered. The judgment question — which of those numbers it should actually take — has not, and it is the more important of the two.
The maximum is a ceiling. It was never a recommendation.
Everything above answers what a lender will extend. The better question is what the business should carry, and the two numbers are not the same. Debt sized to the maximum works only if the plan works; the plan that survives a soft year is the one with room in it. The working discipline is headroom: at least 25 to 30% between your base-case forecast and every covenant in the package, so that a bad quarter is a bad quarter rather than a default, and a conversation with your lender happens from strength rather than from breach. A borrower at maximum leverage has spent that headroom before anything has gone wrong — and has also spent the capacity that would have funded the next acquisition, the fit-out or the contract mobilisation, because a lender’s ceiling does not move simply because you have already reached it.
Sometimes the honest answer is that the capacity is less than the plan hoped, and the right response is to change the funding mix rather than to force the debt. The gap between what the lenses support and what the deal needs can be closed from several directions, each with its own price. Equity — from the owners, or from an outside investor — costs ownership but carries no covenant and no coupon. On an acquisition, a vendor loan note or deferred consideration leaves part of the price with the seller, payable out of the cash flows the deal itself generates, and aligns the seller with the business they are handing over. And staging the plan — the first acquisition now, the second from the deleveraged balance sheet in eighteen months — often raises the same total at a fraction of the risk of stretching to do everything at once. None of these is a failure of the financing; each is what disciplined structuring looks like when the arithmetic is respected rather than argued with.
The pattern to avoid has a familiar shape: capacity treated as a target, raised in full at the top of a good year, against covenants set tight to the forecast. Every element of it is individually defensible and the combination is how good businesses end up restructuring in mediocre years. The borrower who takes less than the maximum keeps the option to come back for more; the one who takes it all hands the option to the lender.
We will run the three lenses on your numbers.
If you want a straight read on your own capacity — what a bank would extend, what a fund would stretch to, what the balance sheet would support, and which of those numbers you should actually take — a first conversation is confidential and costs nothing. We size the lenses on your figures before any lender does, which is also where we tell you plainly if the answer is less than you hoped, and what the funding mix that respects the arithmetic looks like. See how a mandate runs in how we work, or the full range of what we advise on in our services.