A loan covenant is a promise the lender can act on.
A covenant is a contractual promise inside your facility agreement. Break one and the lender gains a defined right to respond, up to and including calling the loan, though it rarely comes to that. Covenants come in two families. Financial maintenance covenants are the ratios (leverage, interest cover, debt-service cover) that your business must keep passing at every test date, and they are what most people mean by the word. Undertakings are the do and do-not promises around them: deliver accounts on time, keep the insurance in place, do not take on more debt or grant fresh security without consent. This guide is about the financial covenants, because they are the ones a borrower can shape and the ones that bite. It sets out what each covenant tests and its typical level, the difference between maintenance and incurrence, how headroom is set and why too little of it is the real danger, cov-lite and springing structures, how quarterly testing works, and what is negotiable before you sign.
Written for the borrower’s side of the table. This is general guidance, not advice on your specific facility. It is the first stop in the covenant journey, ahead of negotiating the term sheet and, if trading dips, handling a covenant breach. It deepens the shorter answers in our working guide.
Covenants are the lender’s early warning and its seat at the table.
A lender advancing against cash flow is repaid out of earnings it cannot see and does not control. Covenants convert that exposure into a set of measurable promises tested on a schedule, so deterioration shows up in the numbers a quarter or two before it shows up in a missed payment. That is the first job: information. A compliance certificate that reports leverage drifting toward its limit tells the lender the credit is softening while there is still time to talk about it.
The second job is control. Because a covenant breach is an event of default, the ratios give the lender a defined right to act (to reprice, to tighten terms, to ask for more information, in the last resort to enforce security) before value leaks out of a weakening business. Set sensibly, covenants are not a trap sprung for its own sake; they mostly do their work silently, and a well-run borrower can go the life of a facility without one ever being tested in anger. The point of understanding them is to make sure the levels and the definitions are set so that ordinary trading variance never trips a wire that was meant for real distress.
Alongside the financial ratios sit the undertakings: positive ones (deliver management accounts and audited statements to a timetable, maintain insurance, keep the security in place) and negative ones (do not incur further debt, grant security or pay dividends beyond agreed limits without consent). Those matter, but they are promises about conduct rather than performance. The covenants that respond to how the business actually trades are the financial ones, and they are what the rest of this guide is about.
Leverage, interest cover and cash cover, tested every quarter.
A £3m to £15m facility typically carries two to four maintenance covenants, tested quarterly on a trailing twelve-month basis. Each one tests a different thing, and the levels vary with the structure and the lender category. What follows is what each measures and where the levels usually sit; your own documents set the numbers that bind.
Leverage · net debt to EBITDA
The headline gearing test: net debt divided by trailing twelve-month EBITDA. Bank senior debt commonly sits around 2.5 to 3.5 times; a unitranche or private-credit structure runs higher, roughly 4 to 4.5 times, with total leverage including any junior debt reaching the upper end. An owner-managed or conservatively-geared senior deal is often nearer 1.5 to 3 times. Because it is measured against a debt balance that barely moves quarter to quarter, a dip in earnings moves the ratio sharply, which is why leverage is usually the covenant that trips first. Why the bands differ by lender is in our comparison of unitranche and bank senior debt, and how leverage caps the size of the raise is in how much your business can borrow.
Interest cover · EBITDA to interest
Tests whether earnings comfortably cover the cost of the debt, the multiple set to the deal and tightening as leverage rises. It is the covenant the rate environment moves directly. Floating-rate facilities are priced at a low single-digit margin 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. As base rates rose the interest line rose with them, so interest cover compressed even where trading held, which is why the covenant deserves as much attention as leverage when you price the deal. The full pricing stack is in our guide to the all-in cost of raising debt.
Debt-service cover · DSCR and cashflow cover
A debt-service cover covenant measures earnings, or cash, against the full cost of the debt: interest plus scheduled amortisation. It is common on amortising bank debt and asset-backed structures, and it is usually set a little above 1.0 times, so cash comfortably covers the whole debt service rather than the interest alone. A cashflow-cover covenant is the stricter cousin, measuring free cash flow after tax and capital spending against debt service, which is a harder test for a business that invests heavily.
The capex limit
Not a ratio but a ceiling: a cap on annual capital expenditure, protecting the cash the lender is counting on to service the debt. It usually allows a modest carry-forward of unused headroom and a permitted overspend funded from new equity. It is the covenant borrowers most often forget they have, until a growth year runs into it.
Maintenance covenants test you every quarter; incurrence covenants test only when you act.
A maintenance covenant is tested at every measurement date whether or not you have done anything. You pass it each quarter or you are in breach, so a passive drift offside (a soft year, a lost contract) is itself the breach. An incurrence covenant works the other way: it is tested only when you take a specified action, such as raising more debt, paying a dividend or making an acquisition. You must show the ratio would still sit within an agreed level after the action, but if you simply do nothing, a ratio that has drifted is not a breach. The distinction decides how much room you have to trade through a bad patch without a conversation with the lender.
At £3m to £15m, expect maintenance covenants. Fully incurrence-based, covenant-lite packages are a large-cap and broadly-syndicated feature, seen on some of the larger unitranche deals but rarely below them, and their borrower-friendliness is precisely why lenders reserve them for the strongest credits. What a lower-mid-market borrower does meet is incurrence logic living inside the undertakings: a permitted-acquisitions or permitted-debt basket you can only draw on if pro-forma leverage stays under a set level. That is an incurrence test in all but name, and the level it is set at is worth negotiating.
The practical read-across: maintenance covenants are what you manage quarter by quarter, and they are the ones a breach guide is written about. The incurrence-style baskets are what you plan around when you want to do something (buy a competitor, take a dividend, add a tranche of debt), and they are set at term-sheet stage alongside the maintenance levels.
Headroom is the gap between your forecast and the covenant, and too little is the real risk.
Covenant headroom is the cushion between the level your base-case plan produces and the level the covenant is set at. Customarily it is 25 to 30%, and it is sized by running the downside case, not the plan. You take the forecast, stress it with the things that realistically go wrong (a revenue miss, a margin squeeze, a delayed contract, a slow debtor), and set the covenant so a credible bad quarter still passes. Headroom that only survives the base case is not headroom; it is a covenant set on top of the plan, waiting for the first ordinary miss.
A worked illustration, at illustrative levels rather than a quote: a business forecasting leverage of 2.8 times, with the covenant set at 3.5 times, carries roughly 25% of headroom (3.5 divided by 2.8). Whether that is enough is not a matter of the percentage; it is whether a realistic downside keeps leverage under 3.5 times. If the stressed case pushes it to 3.6, the 25% is decorative. Headroom is only ever measured against your own honest downside, which is why the forecast that sizes it is the real work.
Too little headroom is the risk that actually costs money. A covenant set right on top of the plan turns normal trading variance into an event of default, a waiver fee and a business-support conversation that a slightly looser level would have avoided entirely. Borrowers fixate on the headline ratio; what protects them is the distance between that ratio and their realistic downside. What happens when the headroom runs out (the cure windows, the price of a waiver, who inside the bank you end up dealing with) is set out in our guide to a covenant breach.
In fund and unitranche structures the covenant may only spring when the revolver is drawn.
A covenant-loose or covenant-lite package carries fewer maintenance covenants than a fully-covenanted bank deal, often a single leverage covenant in place of a suite of three or four. The springing covenant takes it a step further: the financial covenant is tested only when the revolving facility is drawn past a set threshold, often something like 40% of the revolver’s size. Leave the revolver undrawn or lightly drawn and there is no maintenance test to breach in an ordinary quarter. These structures live at the fund end of the market, on unitranche and other private-credit deals, and our comparison of unitranche and bank senior debt sets out where they fit.
The reprieve is smaller than it looks. Where a maintenance covenant is removed, the lender’s protection does not vanish; it migrates into the event-of-default and information provisions, and a fund lender watches the monthly numbers closely regardless. A missing covenant is not a missing lender. What changes is where the conversation starts. Without a quarterly test, a softening credit is flagged later and on the lender’s reading of the monthly pack rather than on a hard ratio, which is not always in the borrower’s favour, because a conversation that starts late tends to start from a weaker position.
For a £3m to £15m borrower the practical point is expectations. Bank senior debt at this size is almost always fully covenanted with maintenance tests, so the cov-lite and springing conversation belongs to the fund-debt lanes, and it is one more axis on which the lender categories differ rather than a free option every borrower can ask for.
Quarterly, on trailing twelve-month numbers, certified by you.
Financial covenants are tested as at each quarter-end, almost always on a trailing twelve-month basis so that a single strong or weak quarter is smoothed across the year. After the quarter-end you have a delivery window, commonly thirty to sixty days, to submit a compliance certificate: a signed statement, typically from a director or the finance director, that sets out each covenant, its required level, the actual figure and the calculation behind it, and confirms compliance or reports a breach.
The timing point borrowers routinely miss: the covenant is tested on the test date, not on the day you file the certificate. The moment the quarter’s management numbers are in hand, well before the certificate is due, you already know whether you passed and by how much. That window between the test date and the filing is not a grace period; it is your time to decide how to report and whether to open a conversation with the lender before the document lands rather than after. Modelling the covenant forward, rather than waiting for the certificate to tell you, is the single most useful discipline in the whole process.
The calculation is not mechanical, because the inputs are defined terms. EBITDA, net debt, how leases and shareholder loans are treated, and what add-backs are permitted are all set by the definitions in the agreement, and they can move a ratio materially from the figure your statutory accounts would suggest. That is why the definitions matter as much as the levels, and it is where the negotiation in the next section does most of its work. If a certificate is going to report a trip, the moves that keep the most options open are in our guide to a covenant breach.
The definitions, the headroom, the equity cure and the add-backs, not just the ratio.
Covenants are set at term-sheet stage, and most of the package is negotiable, but the negotiation that matters runs deeper than the headline number. The lender proposes a package calibrated to protect itself; the borrower’s job is to trade the levers that cost the lender little and buy real operating room. Four of them do most of the work.
Definitions and add-backs
How EBITDA and net debt are defined decides what the ratio actually measures, so a clear, reasonable add-back list (genuine exceptional items, one-off costs, defined run-rate benefits) can matter more than the ratio itself. This is the most technical lever and the one where borrower-side advice pays for itself.
Headroom
Argue the levels off your realistic downside case rather than the base plan, with the customary 25 to 30% cushion. A covenant a little looser than the lender first offers is the cheapest insurance in the deal.
The equity cure
The right for shareholders to inject equity, or subordinated debt, to fix a ratio for a test, capped over the life of the loan (commonly three or four uses in total). It is now near-universal in leveraged facilities and worth securing even if you never expect to use it.
The number and mix of covenants
Fewer, better-defined covenants beat a long list of tight ones. Whether a lighter or springing structure is on the table depends on the lender category and the strength of the credit. The full playbook for trading these against fees and margin is in our guide to negotiating a debt term sheet.
We will set the covenants to your downside, not the lender’s.
If you are agreeing a facility, or living with covenants that feel tight, a first conversation is confidential and costs nothing. We model the covenants forward on your own numbers, size the headroom off a real downside case, and negotiate the definitions, add-backs and equity cure that keep ordinary trading variance from becoming a default. See how a mandate runs in how we work, or the full range of what we advise on in our services.