You Didn't Deleverage. You Grew Into It.

A distributor I work with went from 4.1x leverage to 3.2x over two years without making a single unscheduled principal payment. Same debt. Bigger denominator. Their banker congratulated them on the discipline.

That's the move most growing companies are running without naming it. You don't pay debt down. You grow past it.

The denominator stopped cooperating

KBRA published its Q2 2026 middle market compendium on July 28, covering 2,785 borrowers and roughly $1.2 trillion of direct lending debt. Median EBITDA growth dropped to 24% from 27% in Q1. Largest quarter-over-quarter decline in the series.

Now look at what didn't move. Median gross leverage held at 6.1x. Median interest coverage held at 1.6x. Both have been flat for two years while EBITDA compounded at better than 25% a year, which tells you something specific: all that growth wasn't buying headroom. It was supporting more debt. The share of borrowers with improving coverage ratios plateaued after two straight years of gains, and KBRA's default monitor hit a record 92 borrowers and $30 billion.

You're not a 6x-levered sponsor portfolio company. Fine. Your lender still reads this research, and your covenant still runs the same arithmetic on a smaller base.

Run the test at flat

An afternoon of work, and almost nobody does it.

Take your covenant definition exactly as written, every permitted adjustment included. Freeze EBITDA at the trailing twelve months. Zero growth. Then layer in what's already committed for the next four quarters. Scheduled amortization. The equipment note funding in Q4. Your earnout payment. Higher pricing on the floating tranche if your grid steps up.

Test each quarter separately.

Clear all four at flat and you have real headroom. Clear Q1 and Q2 and fail Q3, and you don't have a covenant. You have a growth requirement your lender wrote into the credit agreement and you signed.

Growth that only shows up in the denominator isn't deleveraging. It's a bet you renew every quarter, and nobody writes down the odds.

The distributor failed at flat in quarter three. Nothing wrong with the business. Their fixed charge coverage test needed about 9% EBITDA growth to absorb the equipment note, the plan called for 14%, and no one had ever run it at zero. A 5% cushion looks generous until you notice it's the only thing standing between you and a technical default.

Failing early is cheap

Amendments are priced on surprise. Walking into your banker's office in August with a flat-case model, a specific ask, and two quarters of runway gets you a step-down or a cure right for a fee and some documentation. Walking in after a missed test in November gets you repricing, tighter reporting, and a story that follows you into renewal.

Be specific about what you want. Size the equipment note smaller. Push the capex one quarter. Trade a slightly higher rate for a covenant holiday you can actually hit. Bankers have all of these in a drawer, and they hand them out to borrowers who showed up before the breach.

Growth covered for you for three years. Nobody sends a notice when it stops.