Debt service coverage ratio determines lender pricing, the covenant structure a facility carries, and the refinancing capacity a business holds. The formula is simple. The discipline behind the inputs is not, and that gap is where operators surrender terms.
Across the $20M to $100M operator tier, 50 to 65 percent calculate DSCR with inputs a lender will reject. The recalculation costs 25 to 75 basis points in pricing and leaves covenant cushions tighter than they appear. This article walks the formula, the institutional definition of each input, and the discipline that defends the terms a coverage ratio is meant to earn.
DSCR equals cash available for debt service divided by total debt service. A ratio of 1.00x means the business produces exactly enough cash to meet its obligations, with no cushion. A ratio of 1.25x means it produces 25 percent more cash than it needs to cover scheduled debt service. The arithmetic is one division. The complexity sits entirely inside the two inputs, because cash available for debt service is not EBITDA and total debt service is not interest alone. The simplicity of the formula obscures the input complexity, and the input complexity is exactly what a lender underwrites. An operator who divides a generous numerator by an incomplete denominator produces a number that looks like coverage and reads, to a lender, like a calculation that has not been disciplined. The sections that follow define each input the way a credit committee defines it.
The institutional reading of the debt service coverage ratio, from cash available for debt service through total debt service, the formula, and the lender minimum.
Start from EBITDA and subtract unfinanced capex, cash taxes, and cash interest to reach the cash a lender treats as available.
Sum scheduled principal, cash interest, and capital lease payments due over the period.
Divide cash available for debt service by total debt service to produce the ratio.
Test the ratio against the minimum for the relevant loan type to read covenant cushion and pricing capacity.
Cash available for debt service, often written CADS, begins at EBITDA and removes the cash the business must spend before it can service debt. The institutional definition is EBITDA minus unfinanced capex minus cash taxes minus cash interest. EBITDA overstates the cash a lender can rely on because it ignores the capital the business must reinvest, the taxes it must pay, and the interest already committed. CADS corrects for all three.
In this example, the business reports 10,000 of EBITDA but only 5,500 is available to service debt once reinvestment, taxes, and committed interest are removed. An operator who runs the coverage ratio off the 10,000 figure starts the conversation 45 percent too high. The distinction between unfinanced capex and total capex matters: capex funded by a separate equipment line is excluded from this deduction because its financing already sits in total debt service. Capex funded from operating cash, the unfinanced portion, is what reduces CADS.
Total debt service is scheduled principal plus cash interest plus capital lease payments over the measurement period. The error most operators make is treating debt service as interest alone. A lender reads the full obligation: the principal amortization due, the cash interest paid, and the capital lease payments that function as debt even when they are recorded as lease expense. Each component is a cash outflow the business must meet, so each belongs in the denominator.
With cash available for debt service of 5,500 and total debt service of 4,200, the DSCR is 5,500 divided by 4,200, or 1.31x. That clears the 1.25x senior term loan minimum with a modest cushion. Had the operator omitted the 400 of capital lease payments, the denominator would have read 3,800 and the ratio would have shown 1.45x, a figure the lender will not accept. The recalculated 1.31x is the number that survives underwriting, and the 14 basis points of apparent cushion the operator thought it had was never there.
There is no single good DSCR. The minimum is set by where the lender sits in the capital structure and how the position is protected. The table below states the working references, and the gauge underneath lets you move the minimum marker across each instrument.
| Loan Type | DSCR Minimum | How a Lender Reads It |
|---|---|---|
| SBA 7(a) | 1.15x | A program minimum, thin by design because the federal guarantee absorbs part of the loss risk. |
| Senior Term Loan (LMM) | 1.25x to 1.50x | The working reference for established operators. The cushion widens with cyclicality and concentration. |
| Asset-Based Loan | 1.10x to 1.20x | Lower because the borrowing base carries most of the protection, not the coverage ratio. |
| Mezzanine | 1.05x to 1.15x | The thinnest on a total-debt-service basis, priced for subordination behind senior debt. |
| Unitranche | 1.20x to 1.35x | A blended senior and subordinated read held by a single lender, which tightens the input definitions. |
1.15x
The 7(a) program sets a 1.15x floor as a program requirement, not a negotiated covenant. The cushion is thin by design because the guarantee absorbs part of the risk, so the lender reads the inputs closely before relying on the ratio.
The minimum is not arbitrary. It reflects where the lender sits in the capital structure, how much collateral protects the position, and how the facility is priced. A lower minimum on a subordinated or asset-backed instrument is purchased with either yield or collateral, not generosity.
A credit committee does not accept the operator's inputs at face value. It substitutes six adjustments, each of which can move the ratio down. Understanding them in advance is the difference between presenting a number that holds and presenting one that gets marked down in front of the committee.
A lender will strip out pro-forma synergies and projected savings that are not yet evidenced in the trailing statements. The DSCR is read on what the business has produced, not on what it expects to produce.
Annualized run-rate revenue from a strong recent quarter is discounted to the trailing twelve-month figure unless contracted backlog supports it. Run-rate optimism is the most common source of a lender markdown.
Add-backs that the operator claims are normalized are re-tested against the non-recurring, non-operational, and transferable standard. Rejected add-backs lower the numerator and the coverage ratio with it.
Where reported maintenance capex looks suppressed, the lender imposes a capex floor based on the asset base and depreciation, raising the deduction in cash available for debt service.
The lender substitutes its own cash tax assumption rather than accepting a temporarily low effective rate, which can move the numerator by a full quarter of taxes.
Operating leases reclassified as capital leases, and capital lease payments the operator excluded, are pulled into total debt service, raising the denominator.
Each of these adjustments is read alongside the reliability of the numbers behind it. A lender that doubts the governance and reporting discipline of a business will apply the adjustments more aggressively, which is the subject of why lenders read governance before covenants. The defense is to run the substitutions before the lender does, which is what the Lender Readiness Check is built to do.
Each error inflates the ratio the operator presents. The lender corrects it, the corrected ratio is lower, and the gap between the two is priced. The basis-point ranges below are the typical cost of the markdown on a lower-middle-market facility.
Treating EBITDA as cash available for debt service ignores capex, cash taxes, and cash interest. The overstated ratio is corrected in underwriting, and the correction reads as a credibility problem worth 25 to 50 basis points.
Leaving capital lease obligations out of total debt service understates the denominator. Lenders pull them back in, the ratio falls, and the gap between the operator figure and the underwritten figure costs 20 to 40 basis points.
Add-backs without documentation inflate the numerator. When 30 to 50 percent of claimed add-backs fail review, the recalculated DSCR can drop below the covenant and reprice the facility by 30 to 60 basis points.
Annualizing a strong quarter overstates cash available for debt service. The lender resets to the trailing twelve months, and the difference is priced at 15 to 35 basis points plus a tighter covenant.
Reporting suppressed maintenance capex flatters the numerator. The lender substitutes a capex floor tied to the asset base, the ratio falls, and the adjustment carries a 20 to 45 basis point cost.
A defensible DSCR is not the highest number an operator can produce. It is the number that survives the inputs a lender substitutes. The Lender Readiness Check tests cash available for debt service and total debt service against the institutional definitions, applies the pro-forma, run-rate, normalization, capex floor, cash tax, and capital lease adjustments, and reports the coverage ratio a credit committee would actually underwrite. Where the gap between the presented ratio and the underwritten ratio is wide, the cost is read in pricing and covenant headroom.
The coverage ratio is one read inside a larger structure. The Capital Structure Stress Test models how DSCR moves under rate, revenue, and capex shocks across the full debt stack, and the Capital Readiness Scorecard places the coverage read inside the seven dimensions a capital partner weighs. The input terms are defined in the glossary, and the infrastructure that makes the numbers reliable is built through lender and investor readiness.
DSCR is a simple division built on inputs a credit committee reads closely. Run the substitutions before the lender does, defend the ratio, and protect the pricing and covenant cushion it is meant to earn.