Covenants That Still Matter in Private Credit

Here is the order, before any argument for it. At the top sits a financial test measured every quarter. Next comes the flow of monthly information. Then a tight fence around the collateral, an honest definition of earnings, a cap on cash leaving for the owners, a short leash on equity cures, and, last, a rule that sale proceeds repay the loan. The ranking is by one measure only: how much warning each term buys a lender before the damage is done. Terms have loosened in private credit for a long stretch now, and every negotiation involves giving something up. Anyone negotiating who knows the order knows what to trade away first and what to refuse.

First, a maintenance test that tests something

A maintenance covenant is measured on a schedule whether or not the borrower does anything. An incurrence covenant only bites when the borrower takes an action, such as adding debt. The difference sounds technical and is not. Under an incurrence regime a company can drift from healthy to impaired without ever breaching a term, because it never asked permission for anything. The lender learns of the trouble when the interest payment is missed, which is the latest possible moment.

A quarterly test of cash flow against fixed charges, or of total debt against earnings, turns a slow decline into a dated event. Where the test is set counts for as much as whether it exists. A test set so loose that it trips only after the equity is gone is a maintenance covenant in name and an incurrence covenant in practice. I want the threshold placed where the downside case crosses it while the business still has value worth protecting.

Second, reporting that arrives while it is still useful

Monthly financial statements, a covenant certificate each quarter with the arithmetic shown, and an annual budget delivered before the year begins. None of it is glamorous, and borrowers push to trim it to quarterly packages delivered late. That trade is worse than it looks. Numbers that reach a lender two months after the quarter ends describe a company that no longer exists.

The request I care most about is the arithmetic itself. A certificate that states the ratio without showing how it was built invites creative assembly. A certificate that walks from reported results to the tested figure, line by line, lets the lender see which adjustments are carrying the result.

The annual budget earns its place for a different reason. It is the company telling the lender, in advance, what a normal year looks like. When the monthly results begin to separate from the budget, the gap is visible long before any ratio moves, and a lender reading both can ask a pointed question in the second month of a problem instead of the sixth. Borrowers often treat the budget as a formality. I treat it as the baseline every later conversation will be measured against.

Third, a fence around the collateral

The worst surprises in recent loan documents have not come from missed payments. They have come from assets moving. Broad permissions to invest in affiliates, to designate subsidiaries outside the lending group, or to transfer intellectual property can let a sponsor relocate the valuable parts of a company beyond the lender’s reach and then borrow against them from someone new.

The protection is a set of narrow baskets and a flat prohibition on moving material assets, especially brands, patents and key contracts, outside the group that guarantees the loan. This covenant rarely does anything in a good year. In a bad year it decides whether the collateral the lender underwrote is still there to recover.

Fourth, an honest definition of earnings

Every financial test is only as strict as the earnings figure inside it. Modern definitions permit addbacks for savings management expects but has not realized, for restructuring charges that recur annually, and for expected gains from acquisitions that have not closed. Each addback inflates the earnings figure, and a generous enough list makes any ratio pass.

I read the definition of adjusted earnings before I read the covenant levels. A cap on projected savings, a time limit for realizing them, and a requirement that one-time charges be genuinely one-time do more for a lender than tightening the headline ratio. A strict test built on a loose definition is theater.

The definition also travels. It appears in the pricing grid, in the baskets for new debt, and in the permission to pay distributions. Loosening it once loosens every term that refers to it, which is why a lender should fight over the definition early and harder than over any single number.

Fifth, a cap on cash leaving for the owners

Restricted payment covenants limit dividends, distributions and repurchases of equity while the loan is outstanding. Their purpose is simple. Cash that belongs in the business, or in paying down the loan, should not leave for the sponsor while the credit is still exposed.

Borrowers ask for builder baskets that grow with retained earnings and for permission to pay distributions whenever a ratio is met. Some flexibility is fair for a company paying down debt on schedule. What I resist is permission that can be used once, large, just before conditions turn, because a single dividend recapitalization can undo the margin of safety a lender spent the whole underwriting constructing.

Sixth, a short leash on equity cures

An equity cure lets the sponsor fix a covenant breach by contributing cash, which the documents then count as though it were earnings. Used sparingly, it is reasonable. The sponsor is putting money at risk, and that is the behavior a lender wants in a weak quarter.

Used freely, it empties the maintenance test of meaning. A sponsor allowed to cure in consecutive quarters, with no requirement that the cash repay debt, can keep a failing company technically inside its tests while the underlying business deteriorates. The limits worth asking for are few cures across the term, none back to back, and cure money applied to the balance.

A cure that must be applied to the loan changes the sponsor’s arithmetic. The cash no longer props up a ratio for a quarter and then disappears into operations. It reduces what is owed, which leaves the lender better protected if the next quarter disappoints as well.

Seventh, sale proceeds that repay the loan

When a borrower sells a division or a significant asset, the proceeds should reduce the loan unless they are reinvested in the business within a defined window. Without this term, a company can shrink the collateral pool and keep the cash, leaving the lender with the same debt against a smaller business.

It ranks last because it only matters when something is sold, and asset sales are visible events the lender will hear about. It still belongs on the list. Asset sale sweeps are a common casualty of a competitive process, and giving one up quietly is how a lender discovers, years later, that the company it lent to has been sold off in pieces.

The right covenant is the one that forces a conversation while there is still something to discuss.

The order above assumes a lender can negotiate for several of these. Often the room allows fewer. If I were permitted to keep a single covenant and surrender every other protection on the list, I would keep the quarterly maintenance test, set tight and built on an honest earnings figure. Every other term guards a particular door. The maintenance test is the only one that sets a date on the calendar when the lender and the owner must look at the same number together, and in a weakening credit that shared date is worth more than any clause written to handle the fallout afterward.