Debt covenant compliance that sees the step-down coming, cites every add-back and leaves the decision with the treasurer
Every covenant projected quarters ahead, so a step-down never becomes a default.
See one case, screen by screen ↓Debt Covenant Compliance is debt covenant compliance software for treasury that tracks, calculates and projects covenants across a borrower’s credit agreements. Seven agents read every agreement and amendment, compute each ratio as that agreement defines it with every add-back cited, recount it independently, project it six quarters forward against the step-downs, draft the compliance certificates and track every reporting duty. The treasurer decides, the CFO signs and a person delivers to the lenders.
How to choose AI solutions for back-office and enterprise work →
Updated
A borrower with several credit agreements owes each of its lenders a number: net leverage, interest coverage, basket usage, net worth, each calculated the way that agreement defines it and certified by a named Responsible Officer within 45 or 60 days of each quarter end. In most treasuries a debt analyst rebuilds those calculations in a spreadsheet, from definitions that amendments have changed, with the restructuring, transaction and FX items sitting in ordinary ledger accounts. One EBITDA gets used for agreements that each define it differently, and add-back caps go untested.
The real risk is the step-down. Projections are done once a year for the budget, so a leverage limit that tightens from 4.00x to 3.50x is looked at when the quarter is already closing. A missed step-down without a waiver is an Event of Default, with cross-default into the other agreements; a waiver asked for after the breach costs a consent fee, a margin step-up and both sides’ legal fees, at the moment the lenders hold the leverage.
Estimated split of a debt analyst’s and treasurer’s working days for one quarter across four credit agreements, by hand and with the solution.
Seven agents read the agreements, compute and recount every ratio, project the step-downs and draft the certificates; the treasurer decides and the CFO signs.
Tuesday 3 November 2026. Kestrel & Vane Group has four credit agreements and eleven covenants. Its $2.75B Credit Agreement, agented by Halden Northern Bank, has a leverage limit that steps down next March. Here is the morning, screen by screen, as Priya Raman, the Group Treasurer, works through it with the agents.
Overnight the agents closed Q3 under all four agreements and ran every covenant six quarters forward. Her morning opens on four numbers: 11 of 11 covenants in compliance at Q3, one projected breach (3.62x against 3.50x in Q1 ’27), four reporting duties in the next 30 days with one late, and four things that need her, one at a time. Below, what each agent did since Monday 07:00, every claim with its source.
“All eleven covenants hold today; one does not hold after its step-down yet. 4 things need you.”
Net leverage under §7.11(a) is drawn as a line: four quarters actual, six projected from the FP&A forecast of 2 Nov, against a limit that steps from 4.00x to 3.50x on 31 Mar 2027 and to 3.25x on 31 Dec 2027. Q3 ’26 is 3.48x. The forecast books $42.0M of plant consolidation costs to cost of sales across Q4 ’26 and Q1 ’27, taking four-quarter EBITDA to $682.0M against net debt of $2,468.8M. Beside the line, the agreement’s own definitions, with the words that matter highlighted.
“On 31 Mar 2027 the limit steps down to 3.50x. The forecast puts leverage at 3.62x: a breach, five months away.”
The projection agent asked whether any of the $42.0M can be added back. The ratio calculator answers yes: Amendment No. 2 put plant closures and consolidations next to restructuring charges in §1.01 (a)(vii)(B), for every test from 31 Dec 2026. The cap is used to $60.9M of $102.3M, leaving $41.4M, and Q1 ’27 moves from 3.62x to 3.41x, 0.09x of room. Without seeing that work, the independent checker rebuilds EBITDA from 41 ledger accounts and debt from the treasury system: 3.41x, difference 0.00x.
Checker: “One caution: asset write-downs ($11.9M) are inside the clause only if they relate to the closure; the ledger tags them to PCX-26.”
Every number in the thread carries a citation. Priya opens the one on the clause: Amendment No. 2 §2.1(c), signed on 11 Feb 2026 by the Borrower, the Administrative Agent and the Required Lenders. The inserted words are highlighted, and Section 4 makes them apply to any period ending on or after 31 December 2026, so the Q1 ’27 test is covered. She can open it in the agreement or pin it to the evidence.
“(B) costs, charges and expenses relating to plant closures, plant consolidations and the relocation of production lines, including severance, lease termination and asset write-downs”
Accepting needs a reason code (AC-01, permitted by the agreement, clause cited) and a tick that she has read the cited clauses and the ledger extracts. It is signed as Priya and recorded in the audit trail. A minute later the projection agent re-runs six quarters. The tightest covenant is now the notes, 3.66x against 3.75x, because the Note Purchase Agreement’s EBITDA has no plant-closure add-back and caps restructuring at 10%. In the downside the credit agreement still reaches 3.58x, so a covenant-reset request is prepared and held as a draft.
“Nothing goes to the lenders unless you and Legal release it.”
The Q3 calculation runs from Consolidated Net Income of $268.4M to Consolidated EBITDA of $709.4M in nine lines, each with its clause, ledger account, read confidence and the checker’s tick: restructuring of $18.9M under (a)(vii)(A) within the cap, Ardent transaction costs of $7.2M under (a)(viii) from GL 6870, less the $6.8M gain on the Tilbury warehouse. Tomás Ferreira confirmed the ledger close. The caps show what is used, and the run-rate savings add-back sits unused until the CFO certifies it.
“Checker agreed on 2 Nov 07:26 · difference 0.00x”
Projections show every covenant in all four agreements six quarters ahead, beside the levers the agreements allow: the plant-closure add-back (accepted), $16.0M of run-rate savings under §1.01 (a)(ix) that the CFO must certify, and an early covenant reset from the Required Lenders, fee estimated at $2.75M–$6.9M. Priya types “What if steel costs rise 8%?”. The agent raises steel 8% from Q1 ’27, passes half on to customers and answers: 3.47x against 3.50x, still inside by 0.03x; the notes 3.72x against 3.75x. It adds the scenario as a tab.
Net more cash against debt: not allowed. “§1.01 ‘Consolidated Net Debt’ caps netted cash at $250M. Holding the $612M of cash changes nothing, so I do not propose it.”
Every duty in the four agreements is listed with its owner and the Business Day rule applied. Four fall in the next 30 days. The Brazil auditor’s limited-review letter is a day late with Rafael Moreira, reminded on 30 Oct and 2 Nov, and Priya can escalate it to João Lima, Finance Director Brazil, with Rafael staying the owner. The obligation watcher saw Ardent Coatings GmbH become a Material Subsidiary at Q3, at 6.1% of EBITDA against the 5% test: a guarantor joinder under §6.12 by Tue 15 Dec, then the notes’ guarantee under §9.7 by Thu 14 Jan, both with Ade Okonjo.
The obligation watcher never decides whether an event is a Default; Legal confirms.
On 28 Oct Claire Ostrowski of Halden Northern Bank asked how the Ardent integration costs are treated in Consolidated EBITDA. The drafted reply says that only fees, costs and expenses of a Permitted Acquisition are added back under §1.01 (a)(viii): $7.2M for the four quarters, from ledger account 6870, project ARD-24. No other Ardent cost is added back, and net leverage is 3.48x against 4.00x. Priya can approve and send, edit or reject it.
“I did not mention the Q1 ’27 projection: it is not asked, and it is your call when to share it.”
The reporting agent drafted the Q3 certificate from the agreement’s Exhibit D: 3.48 to 1.00 against a maximum of 4.00, 6.80 against a minimum of 3.00, baskets at $142M of $250M and $96M of $150M, no Default, and Ardent disclosed as a new Material Subsidiary. The independent checker’s nine checks pass. Lena prepared it, Priya approves it at 08:33 and Michael Grant signs as Responsible Officer at 08:34. Lena uploads it to the agent’s portal at 08:35 and the reporting agent reads back receipt HNB-AGY-26-118734, 13 days before the deadline of Mon 16 Nov.
Sign-off: “nobody approves what they prepared.”
The same covenant quarter, seen by the five people who carry it: what it looked like, and what it looks like now.
Seven agents read the agreements, compute and recount every ratio, project the step-downs and draft the certificates; the treasurer decides and the CFO signs.
Reads every credit agreement, amendment and side letter; extracts covenants, definitions, step-downs, baskets and reporting duties with clause and page; keeps the conformed copy and knows which text governs which test.
Computes each ratio exactly as each agreement defines it, line by line from the ledger, with every add-back cited and every cap tested.
Recomputes every ratio from the ledger without seeing the calculator’s work, must agree within 0.01x, and checks the certificate’s arithmetic.
Runs every covenant six quarters forward from the FP&A forecast under base, downside and asked-for scenarios, finds the first thin quarter and proposes only levers the agreements allow.
Tracks all 23 duties with the Business Day rule, drafts compliance certificates from each agreement’s form, chases and escalates owners, and reads back delivery receipts.
Watches the ledger, board minutes and the entity register for events that trigger duties or use baskets: new debt, dividends, acquisitions, material subsidiaries.
Drafts answers to lender questions with citations, cover notes, and amendment or waiver requests with fee estimates.
Accepts or rejects each add-back; the CFO signs as Responsible Officer. The agents propose; a named person decides.
Ask about any covenant, clause, deadline or lender, or tell the agents what to do. Every answer cites its source, and anything that changes waits for you to accept it.
Where is our tightest covenant over six quarters?
Credit agreement net leverage in Q1 ’27: 3.62x against the new 3.50x limit, a projected breach before the plant-closure add-back the calculator proposed. With it, 3.41x. Every other covenant keeps at least 0.09x.
Why is the plant-closure add-back permitted?
Amendment No. 2 inserted plant closures, plant consolidations and the relocation of production lines, including severance, lease termination and asset write-downs, into clause (a)(vii) of Consolidated EBITDA, for every test from 31 Dec 2026. Project PCX-26’s $42.0M is exactly that, approved by the board on 24 Sep. The 15% cap leaves $41.4M unused.
What if steel costs rise 8%?
I raised steel costs 8% from Q1 ’27 and passed half on to customers, from the forecast drivers. Credit agreement leverage at the step-down: 3.47x against 3.50x, still inside by 0.03x. Notes leverage 3.72x against 3.75x. I added the scenario as a tab.
What do we owe lenders in the next 30 days?
The Q3 compliance certificate and the Brazil certificate with its auditor’s letter, both due Mon 16 Nov, and the notes officer’s certificate by 30 Nov. Statements are already deemed delivered by the 10-Q. The Ardent joinder follows on 15 Dec.
13 screens from the working solution, on its sample data. Pick one to see it large.
What the agents watch for and the rules they keep are settings. A change to a threshold needs a second approver and is logged.
| Setting | Default | Choose from |
|---|---|---|
| Watch when headroom is below | 0.50x, or EBITDA cushion below 10% | Any level you set |
| Alert, then critical, when headroom is below | 0.25x, then 0.10x | Any levels you set |
| Projection horizon | 6 quarters | Any number of quarters |
| Confirm extractions below | 90% confidence | Any confidence level |
| Checker must agree within | 0.01x | Any tolerance |
| Remind owners, then escalate | 10, 5 and 2 Business Days before; escalate after 1 Business Day missed | Any schedule |
| Anything about a waiver or amendment goes through Legal | On | On · Off |
| Share projections with lenders only when the Treasurer decides | On | On · Off |
Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.
“demo” = seen in the working solution, on its sample data · “estimated” = our estimate (the split of a quarter’s covenant working days) · Context, not claims: IAS 1 amendments “Non-current Liabilities with Covenants” (effective 1 Jan 2024), ASC 470-10-45-11 and SEC Release 33-8350 on covenant disclosure. People, companies, lenders and agreements named on this page are fictional: characters and sample data in the working solution.
Debt Covenant Compliance is debt covenant compliance software for treasury that tracks, calculates and projects covenants across a borrower’s credit agreements. Seven agents read every agreement and amendment, compute each ratio as that agreement defines it with every add-back cited, recount it independently, project it six quarters forward against the step-downs, draft the compliance certificates and track every reporting duty. The treasurer decides, the CFO signs and a person delivers to the lenders.
The ratio calculator computes each covenant line by line from the ledger under that agreement’s own definition as in force on the test date, citing the clause and the ledger account for every add-back and testing every cap. A side-by-side view shows how the definitions differ: in the sample quarter the notes count $49.4M less EBITDA than the credit agreement.
Six quarters by default, from your FP&A forecast under base, downside and any what-if you ask for. Each covenant is drawn against its stepped limit, and the projection agent proposes only levers the agreements allow, each with its clause. In the sample, a step-down breach was seen five months before the test date.
Yes. The reporting agent drafts it from each agreement’s own form, with every figure taken from the checked calculation and cited, and the independent checker runs its checks on the schedule. The debt analyst prepares it, the treasurer approves it and the CFO signs as Responsible Officer; a person uploads it and the agent reads back the receipt.
Yes. Every certificate, statement and notice in the agreements is tracked with its owner and the Business Day rule, and owners are reminded and escalated on a schedule. The obligation watcher adds duties when something happens, such as a guarantor joinder when a subsidiary becomes material; Legal confirms them.
No. An add-back is only proposed with a clause, and it counts only when the treasurer accepts it with a reason code. Agents draft replies and requests but never send them; a person approves and sends, and anything about a waiver or amendment goes through Legal.
Yes. The agents read, compute, check, project and draft; people decide. Nobody approves what they prepared, only the Responsible Officer signs a certificate, and every agent step and human decision is sealed in the audit trail.
The Agentic Solution Engine builds and deploys it from your requirements and documents (your credit agreements and amendments, ledger mappings and forecast), and it goes live once every quality gate has passed. We will walk you through it on your own agreements and last quarter’s close first.
Cash seen 13 weeks ahead, every number traced, so less idle cash sits as buffer.
1-min demoHedge AccountingEvery hedge documented on the trade date and tested each quarter, so earnings stay steady.
1-min demoBank Fee AnalysisEvery bank fee checked against the agreed price, every overcharge claimed with proof.
1-min demoWe’ll run Debt Covenant Compliance on your own credit agreements, amendments and last quarter’s close.