What DSCR actually measures
Debt service coverage ratio asks the simplest question in commercial credit: does the cash a business produces cover the debt it has to pay? The formula every lender writes down is the same — cash flow available for debt service divided by total debt service — and it is the least interesting part of the exercise.
What varies, materially, is what goes into the numerator, what counts in the denominator, and which period is measured. Two competent analysts can work from the same statements, apply the same formula, and arrive at different coverage. Both can be defensible. DSCR is not a fact you extract from a financial statement; it is a policy applied to one.
The numerator is a policy choice
Most shops start at net income and add back interest, depreciation, and amortization. Everything after that is judgment: normalizing owner compensation to a market wage, removing non-recurring gains or losses, deducting distributions the owners must take to pay taxes on pass-through income, treating rent paid to an affiliated real estate entity, and deciding whether maintenance capital expenditures come out before coverage or after it.
Each of those adjustments is reasonable. None of them is automatic. An add-back is a defensible position only when the credit policy names it and the file shows where it came from — which statement, which period, which line. The weakest files are not the ones with aggressive add-backs; they are the ones where nobody can reconstruct why an add-back was taken.
What counts as debt service
The denominator is usually described as principal and interest over the measurement period, and it hides more choices than the numerator. Existing debt plus the proposed facility. Capital lease obligations, and increasingly the operating leases that behave like them. Floating-rate debt, which requires a rate assumption someone has to own. Seasonal working capital lines that are drawn and repaid within the year. Shareholder loans, which count or do not count depending on whether they are formally subordinated and whether the subordination is in the file.
Two structures deserve particular care, because they flatter coverage in the near term and can starve it later: interest-only periods, which understate the eventual payment, and balloon maturities, which move the obligation outside the measurement window entirely. A coverage ratio computed on the first year of an interest-only loan is a true number answering the wrong question. Policy should say whether coverage is tested on actual scheduled payments or on a fully amortizing constant, and the memo should say which was used.
The period question is where committees actually disagree
Coverage is always coverage over something: the most recent fiscal year, a trailing twelve months built from interim statements, an interim period annualized, a pro forma that layers the proposed debt onto historical cash flow, or a projection. These produce different ratios by design.
Mature credit policies define the set — which periods are computed on every file, which one governs the decision, and which are shown for context — rather than leaving it to whoever built the workbook. That definition is what makes coverage comparable across a portfolio. Without it, a lender cannot honestly say its DSCR on one file means the same thing as its DSCR on another, and most committee disagreements about the number turn out to be period disagreements wearing a ratio's clothing.
Program lending tightens this further. Guaranteed and mission-driven programs carry their own coverage expectations and their own rules about what may be normalized, and those requirements change over time. A lender running several programs is not running one DSCR definition; it is running several, and it needs to know which one applied to a given file on the date the file was decided.
Global coverage and multi-entity borrowers
Most complex commercial credits are not one entity. There is an operating company, often a real estate holding entity that owns the building, sometimes affiliates that share management, and guarantors with personal obligations. Global DSCR consolidates that picture — and consolidation without elimination rules produces a number that is quietly wrong.
The recurring failure is double counting: rent shows up as an expense at the operating company and as income at the holding entity, and both survive into the global calculation. Management fees between affiliates behave the same way. On the guarantor side, personal cash flow rarely arrives with the rigor applied to business statements, and personal living expenses are the line most often estimated and least often documented. Global coverage is a consolidation exercise, and the eliminations are where files get thin.
What makes a DSCR defensible
An examiner, a credit committee, or an investor performing diligence on a file is rarely asking whether the ratio is high enough. They are asking whether it is reproducible: is the definition written down, do the inputs trace back to a source document and page, is the period labeled, are the eliminations stated, and are analyst overrides recorded with a reason? Consistency across files matters as much as the value on any one of them, because consistency is what makes a portfolio-level statement about credit quality mean anything.
That is also the practical test for software. Coverage computed from a confirmed spread rather than re-keyed into a workbook keeps the ratio tied to its evidence. In CORE, the definitions — which add-backs, which debt, which periods, which entities roll up — live in the program configuration rather than in one analyst's spreadsheet, so a loan program computes coverage the same way on every file it touches. CORE Underwriting shows the inputs behind each ratio back to the source page, and Program Studio holds the policy tests a program applies. The model proposes the calculation; the analyst confirms it, and only confirmed figures carry into policy tests and the memo. Automate the arithmetic and the bookkeeping around it; never automate the judgment about what belongs in the numerator.
Common questions
What is a good DSCR?
There is no universal figure. Minimums are set by the lender's credit policy and, for guaranteed programs, by the program's own requirements. What matters in a file is that the threshold is stated in policy, applied consistently, and measured over a defined period.
What is the difference between DSCR and global DSCR?
Entity-level DSCR measures one borrower's cash flow against its own debt. Global DSCR consolidates affiliated entities and guarantors, which requires elimination rules so intercompany rent, management fees, or distributions are not counted twice.
Which period should DSCR be measured over?
Whichever periods credit policy defines — commonly the most recent fiscal year, a trailing-twelve-month interim, and a pro forma including the proposed debt. The requirement is less about which period wins than that the file states which one governed the decision.
Go deeper
Where coverage is computed in CORE Underwriting →
Program-defined policy tests in Program Studio →