Three coverage tests, one set of cash flows, and a full turn between them.

What does each ratio actually measure?

Interest coverage asks whether earnings cover the coupon. In its common form it is EBITDA divided by cash interest expense for the period. It ignores principal, capital expenditure, taxes, leases and everything else a business has to pay, which makes it the most forgiving of the three and the easiest to compare across companies. It is also the ratio most sensitive to the base rate, because the numerator is annual and the denominator reprices every quarter on a floating-rate facility.

Debt service coverage adds the principal that falls due. Cash flow, sometimes EBITDA and sometimes a defined cash flow available for debt service, is divided by cash interest plus scheduled amortization for the period. This is the standard test in asset-backed, real estate and bank term lending, because those structures amortize. In a bullet unitranche with 1% annual amortization the difference from interest coverage is small. On a bank term loan amortizing at 10% a year it is very large.

Fixed charge coverage is the strictest, and the least standardized. The usual construction takes EBITDA, subtracts unfinanced maintenance capital expenditure, cash taxes and often distributions and rent, then divides by cash interest plus scheduled principal and sometimes rent again in the denominator. It is trying to answer a different question from the other two: not whether earnings cover the debt, but whether what is left of earnings after the business has paid to stay in business covers the debt. That is why it is the test most credit agreements in the lower middle market actually use.

None of these definitions is fixed. Each appears in credit agreements in several forms, and the drafting decides the outcome as much as the arithmetic does. Whether maintenance capital expenditure is defined by the borrower or capped at a percentage of revenue, whether cash taxes are actual or notional, whether rent sits in the numerator or the denominator, whether payment-in-kind interest counts as interest at all: those choices routinely move the ratio by more than a quarter of the business's trading does.

Why can a business pass one test and fail another?

Because they are measuring different obligations out of the same cash. Take a business with $10m of EBITDA and $40m of floating-rate debt at an all-in cost of 10.0%, amortizing at 5% a year, spending $1.2m a year on unfinanced maintenance capital expenditure and paying $0.8m of cash taxes. Its interest coverage is 2.5x, which reads comfortably. Its debt service coverage is 1.7x, which reads adequately. Its fixed charge coverage is 1.33x, which reads close to a covenant floor.

Now move the base rate up by 100 basis points and hold everything else. Interest coverage falls to 2.27x and still looks fine. Debt service coverage falls to 1.56x. Fixed charge coverage falls to 1.25x, which in most lower middle market credit agreements is the level at which the conversation with the lender begins. The business has not changed. One number in the denominator has, and only the strictest of the three tests registers it as material.

That is not a hypothetical direction of travel. The Federal Reserve held its target range at 3.50% to 3.75% on 29 July 2026 by a 9 to 3 vote, with all three dissenters voting to raise, the first time since September 2016 that three policymakers dissented in the same direction (FOMC statement, 29 July 2026). Three-month Term SOFR stood at 3.76% on 4 August 2026 and the forward curve prices it at 4.04% at the end of 2027 (CME and Blue Gamma, 4 August 2026). A 100 basis point move is a stress test rather than a forecast, but the direction the curve prices is up, not down.

Which test is the one that binds?

In a credit agreement, whichever one is drafted with the least headroom, which in the lower middle market is usually fixed charge coverage. Maintenance covenants remain the norm in smaller transactions (First Eagle Investments, March 2026, citing PitchBook LCD as of 28 February 2026), and the fixed charge test is the one that captures amortization, capital expenditure and taxes together.

The distinction matters in a second place borrowers rarely look at until it is urgent. A debt incurrence test typically permits new debt if either a leverage ratio or a fixed charge coverage ratio is satisfied. The leverage test is not sensitive to interest rates. The fixed charge test is (Kricheff, A Pragmatist's Guide to Leveraged Finance). In an environment where base rates are flat to rising, one of those two tests does all the binding, and it is not the one that gets modelled.

The market data says the same thing from the other direction. Median gross leverage across 2,785 middle-market borrowers and more than $1.2 trillion of debt held at 6.1x, unchanged on the quarter, while median interest coverage sat at 1.6x and the share of borrowers with improving coverage plateaued after more than two years of gains (KBRA, Q2 2026 Middle Market Compendium, for the twelve months ended 30 June 2026, published 28 July 2026). The leverage number has barely moved in years. The coverage number is the one that repriced when the base rate did, and it is the one that stopped improving.

What flatters a coverage ratio?

Three things, in ascending order of consequence. Add-backs to EBITDA lift the numerator of all three tests at once, which is why they are the most contested item in any credit negotiation. The Financial Stability Board notes that stripping EBITDA adjustments out of private credit borrowers moves leverage from a reported 5x to 6x range to something closer to 7x (Financial Stability Board, Report on Vulnerabilities in Private Credit, 6 May 2026). Whatever that does to a leverage covenant, it does the same to a coverage covenant in the other direction.

Deferring capital expenditure lifts fixed charge coverage immediately and only fixed charge coverage, because it is the only one of the three that subtracts it. A ratio that improves for two quarters while the asset base ages is not an improving business. Lenders know the pattern and ask for the maintenance capital expenditure schedule, not the reported figure.

And converting cash interest to payment-in-kind lifts every test at once by removing cash interest from the denominator. Payment-in-kind interest was present on 10.6% of direct lending loans and represented 8.9% of total interest income in the first quarter of 2026, while loans that carried no payment-in-kind at close but do today reached 5.9% of all loans, a measure Lincoln itself describes as a shadow default rate. Loan to value on that cohort rose 33.5 points to 76.0% over the year (Lincoln International, as of 31 March 2026). Lincoln has also observed that the market-wide improvement in fixed charge coverage may be partly an artifact of that same shift (Lincoln International, 11 February 2026). A coverage ratio improved by payment-in-kind is a ratio that improved by borrowing more.

What is the market actually covering?

Fixed charge coverage across a large middle-market sample sat at 1.3x in the first quarter of 2026, the third consecutive quarter at that level and up from a 1.1x trough in the first quarter of 2024, while the share of borrowers below 1.0x fell to 19.5% from a 40.9% peak in the second quarter of 2024 (Lincoln International, as of 31 March 2026). Read the second figure carefully. Roughly one borrower in five in that sample was not covering its fixed charges at all, in a market that had been improving for two years.

At the small end the picture is more conservative and the ratios are healthier, because the leverage is lower. One lower middle market lender reported a portfolio with median net senior debt to EBITDA of 2.5x and median EBITDA to senior interest of 3.0x, on an average portfolio company EBITDA of $11.2m (Main Street Capital, Form 8-K, as of 31 March 2026). Another reported new platform deals at 2.8x weighted average senior leverage and 29% weighted average loan to value (Capital Southwest, Form 8-K, for the quarter ended 30 June 2026). Coverage at those levels is not the binding constraint. Leverage capacity is.

Higher up the market the reverse applies, which is the whole point of separating the tests. At the 5.00x to 6.50x total leverage that clears for borrowers above $25m of EBITDA (SPP Capital Partners, Market At A Glance, July 2026), interest alone consumes enough of the earnings that the fixed charge test becomes the constraint on the last turn of debt. The same business, financed two turns lighter, would never touch it.