LHThe Long Hold
← All issues

Leverage rarely kills a company on its own. Covenants and timing do.

How to read a credit agreement the way an operator has to, rather than the way a spreadsheet does.
Capital · September 24, 2026 · 4 min read · Written by Ahmad Khan

Ask most founders how much debt their business carries, and you'll get a number. Two and a half million, or three times EBITDA. Something clean.

That number is rarely what matters. What matters is when it gets tested, what happens if you fail, and how much room you have in between.

I've seen profitable businesses go into default and unprofitable ones sail through. The size of the loan was never the difference.

The number is an average. The covenant is a moment.

You underwrite a business on its annual performance. Revenue does this, margin does that, and cash flow covers the debt service one and a half times over. Comfortable.

But a covenant doesn't test the year. It tests a date, usually the last day of a quarter, and it looks at a trailing figure on that date. If your business has any seasonality at all (and almost every business does), those aren't the same thing.

A company having a perfectly good year can still fail a test in the one quarter where a big renewal slips, a large collection lands three days late, or the annual insurance premium hits in the same month as bonuses.

You don't default because you can't afford the debt. You default because of where the calendar happened to fall.

So the first thing to do with a credit agreement isn't to check the interest rate. Write down the next eight test dates and hold each one up against what you know about your own cash cycle.

The four things to find

Most agreements are long, but only a handful of clauses will ever matter to how you run the business.

The tests, and their dates. Usually leverage and coverage. Check whether they tighten over time. Many step down each year, which means a business that's simply standing still drifts closer to breach without doing anything wrong.

The definition of EBITDA. Not the accounting definition, but the one in this document. It spells out which add-backs are allowed and often caps them. If you report on one definition and get tested on another, you can be badly surprised.

The cure rights. Can a shareholder put in cash to fix a failed test? How many times, and in how many consecutive quarters? An equity cure is often the difference between a bad quarter and a restructuring.

The cash sweep. How much of your surplus cash is forced toward repayment, and at what leverage levels does that step down? This is the clause that quietly decides whether you can fund next year's hires out of your own profits.

Headroom is the real metric

Leverage tells you what happened. Headroom tells you what you can survive.

If you can absorb a thirty percent drop in EBITDA before breaching, you have a structure with room to be wrong. If it's eight percent, you need the next two years to go roughly to plan, and in my experience they rarely do.

Headroom also moves. It shifts with the step-downs, with seasonality, and with any one-off in the trailing period. The tightest point often isn't the next test. It's three or four quarters out, when a step-down lines up with the trailing window rolling off a strong quarter.

THE COVENANT CHECK

  1. Find the tightest test. Map out the next eight quarters, step-downs included. The binding moment is rarely the next one.

  2. Measure your headroom on that date. How far can EBITDA fall before you breach? Put it as a percentage, not a feeling.

  3. Know your cure before you need it. How many times can a shareholder put in cash, and in how many consecutive quarters?

  4. Work out what surplus cash is actually yours. After the sweep, what's left to reinvest?

You don't need a banker to answer any of these, and every one of them changes how you'll run the next twelve months.

What this changes about operating

Timing of spend becomes strategic. The question isn't whether to make the hire but which side of the test date the cost lands on. That isn't gaming anything. It's the same discipline as managing working capital.

Revenue quality matters more than revenue. A big one-off fee flatters the trailing figure for four quarters, then drops out and tightens your leverage at a moment you didn't choose.

Bad news should travel early. The most useful habit with a lender is flagging a problem a quarter before it shows up in a test, with a plan attached. Nobody wants to enforce, but everybody wants to be told.

The one thing worth remembering

People talk about debt as if it were a weight: heavier is worse, lighter is better. I find it more useful to think of it as a schedule you've agreed to keep.

Businesses rarely fail because the weight was too heavy. They fail because they missed a date they hadn't looked at closely enough.

— Ahmad

THIS WEEK'S MOVE

Pull out your credit agreement and write down the next eight test dates. It takes twenty minutes. Don't analyze anything yet.

Thinking about taking on capital?

Talk through the decision with me.

I set aside a small number of 30-minute founder conversations each month. We'll use the time to understand where the business is today, pressure-test one important question, and identify what deserves attention next.

Get the next issue

One operating framework, every Thursday. Free, and you can unsubscribe anytime.

Free. Weekly. Unsubscribe anytime.

← All issues
LHThe Long Hold is written by Ahmad Khan, an investor and operator thinking about technology, ownership, and what it takes to build businesses that last.
© 2026 The Long Hold · generationgp.com