DSCR Formula: Every Variant Lenders Use and When Each One Applies

There is no single DSCR formula. There is a family of them, and the same borrower produces a materially different ratio under each. The general form is cash available for debt service divided by debt service in the same period — but every lender defines both terms differently, and the facility agreement's definition is the only one that binds.

This page runs one borrower's financials through ten variants. The answer ranges from 0.84x to 1.66x on identical accounts. YuSight computes each variant from the spread with 100% of figures cited back to the source document and page, so an analyst can see which line moved the ratio rather than arguing about it.

Key facts

• The OCC's Commercial Loans booklet instructs examiners to "determine, for term loans, if the payment terms are consistent with the type of asset financed and assess whether operating cash flow is sufficient to meet the scheduled amortizing payments" — a coverage test, with no formula prescribed (Comptroller's Handbook, Commercial Loans).
• The Interagency Guidance on Leveraged Lending fixes one input by name: "Cash should not be netted against debt for purposes of this calculation" (SR 13-3 attachment, Federal Reserve). Almost nothing else about coverage arithmetic is fixed anywhere.
• IFRS 16, effective for annual periods beginning on or after 1 January 2019, puts a right-of-use asset and lease liability on the balance sheet for leases over 12 months unless the asset is of low value (IFRS Foundation). A borrower who has adopted it reports higher EBITDA and therefore a higher DSCR than an identical borrower who has not.
• In the worked example below, the same borrower scores 1.66x on an EBITDA basis and 0.95x on a fixed charge basis — a 0.71x spread produced entirely by definition, not by performance.
• Every DSCR floor quoted in this article is lender policy, not regulation. Neither the RBI, the US federal banking agencies nor the CBUAE prescribes a minimum DSCR for general commercial lending.

What is the DSCR formula, in its general form?

DSCR = Cash available for debt service in the period ÷ Debt service falling due in the same period

Both halves are contested.

The numerator can be EBITDA, EBITDA after tax and capex, net operating income, profit after tax with non-cash items added back, or a group cash flow including a guarantor's household. The denominator can be interest only, interest plus scheduled principal, interest plus principal plus rent and distributions, or a stressed constant the lender invents.

Getting from a set of accounts to any of them is a spreading problem before it is a ratio problem — see what financial spreading is and how it works.

The borrower: one set of accounts, ten answers

Northgate Lodging is a 120-key limited-service hotel operating company, owner-managed, with a term loan secured on the property and a ground lease on the site. Year ended 31 March 2026. Figures in currency units — the arithmetic works in any currency.

Unless a variant says otherwise, debt service = cash interest 1,180 + scheduled principal 1,500 = 2,680.

Every variant, side by side

Same accounts. 0.84x to 1.66x. If your credit committee approved at 1.40x and the facility agreement tests on fixed charges, you approved a borrower who was already in breach on the day of drawdown.

How is standard EBITDA DSCR calculated?

DSCR = EBITDA ÷ (Interest + Scheduled principal) = 4,460 ÷ (1,180 + 1,500) = 4,460 ÷ 2,680 = 1.66x

This is the version quoted in credit committee papers, and it is the most generous of the group. EBITDA ignores tax, ignores the capex needed to keep the hotel open, and ignores working capital. It answers "does the business generate enough operating profit to look like it can pay?" — not "will the cash be there?"

Two immediate refinements an analyst should make:

• Accrued or cash interest? Using the full accrued 1,240 rather than cash 1,180 gives 4,460 / 2,740 = 1.63x. The PIK accrual on the shareholder note is real interest that is not cash this year. Which one the covenant uses is a drafting question, covered in covenant testing for DSCR and leverage.
• Which principal? Scheduled amortisation only, or does a balloon maturing inside the test period land in the denominator? A 1.66x borrower with a bullet repayment in month 11 is not a 1.66x borrower.

How does DSCR on CFADS differ?

Cash flow available for debt service strips out everything EBITDA politely ignores.

CFADS = EBITDA − Cash tax − Maintenance capex − Increase in working capital = 4,460 − 518 − 620 − 180 = 3,142 DSCR = 3,142 ÷ 2,680 = 1.17x

The ratio falls by nearly half a turn. Nothing about the business changed; the numerator simply stopped pretending that a hotel can defer soft-goods replacement indefinitely. CFADS is the standard base in project and infrastructure finance and increasingly in asset-heavy corporate lending. Where a borrower is capex-light and tax-sheltered, CFADS and EBITDA converge — which is exactly why a single house convention across a portfolio produces misleading comparisons.

What is the Indian cash-profit DSCR convention?

Indian bank appraisal notes, CMA formats and project reports overwhelmingly use a cash-profit build-up rather than EBITDA:

DSCR = (PAT + Depreciation + Interest on term loan) ÷ (Interest on term loan + Principal repayment) = (1,552 + 1,150 + 1,180) ÷ (1,180 + 1,500) = 3,882 ÷ 2,680 = 1.45x

The gap to the 1.66x EBITDA figure is exactly 578 units, and it is worth decomposing because it catches analysts out:

• 518 — tax. PAT is after tax, and the convention does not add it back, so cash-profit DSCR is a post-tax measure while EBITDA DSCR is pre-tax.
• 60 — the PIK accrual on the shareholder note. The convention adds back interest on term loan, not all interest. Any interest expense outside that definition is deducted in PAT and never restored.

Two further traps in Indian practice:

1. Working capital interest. Interest on cash credit and working capital demand loans is normally excluded from both numerator and denominator, on the reasoning that the working capital facility revolves and is not being amortised. If it is added to one side and not the other, the ratio is nonsense. See drawing power versus ratio covenants in cash credit facilities.
2. Average DSCR over the loan tenor. Indian term-loan appraisals typically quote an average DSCR across the full repayment period alongside a minimum, not a single-year figure. That is the project finance treatment described below, applied to ordinary corporate term lending.

The RBI's (Project Finance) Directions, 2025, issued 19 June 2025 and effective 1 October 2025, require that the "pre-dominant source of repayment... must be from cash flows arising from the project which is being financed" and that the "post DCCO repayment schedule has been realistically designed to factor in the initial cash flows" (RBI). The Directions frame the cash-flow test without prescribing a DSCR formula or a floor — the arithmetic remains lender policy.

How is global DSCR calculated?

Global DSCR consolidates the operating business with the guarantor's household, and sometimes with affiliates. It is the default in US small-business and owner-managed commercial lending.

Global DSCR = (Business cash flow + Guarantor net personal cash flow) ÷ (Business debt service + Personal debt service)

Northgate's owner-manager:

Global DSCR = (4,460 + 420 + 260 − 130 − 540) ÷ (2,680 + 420) = 4,470 ÷ 3,100 = 1.44x

Two double-count traps decide whether this number is honest:

• The owner's salary. Northgate pays the owner 300, and that 300 is already an operating cost inside EBITDA. It is excluded from personal income above. Count it in both places and the global DSCR inflates by roughly 0.10x on these numbers.
• Distributions. The 900 paid to the shareholder is a transfer between the two halves of the same consolidation. Adding it to personal income while leaving business EBITDA untouched counts the same cash twice.

The market-specific mechanics for SBA 7(a) — affiliate treatment, which household expenses count, which SOP applies — are set out in how global DSCR is calculated for SBA 7(a) loans.

What is CRE DSCR, and why is it not the corporate number?

Commercial real estate lenders underwrite the property, not the operating company.

CRE DSCR = Net operating income ÷ Debt service

NOI is built from EBITDA by removing what belongs to the corporate entity and imposing what the lender insists a prudent owner would spend:

CRE DSCR = 3,678 ÷ 2,680 = 1.37x

The management fee is deducted whether or not it is paid. The owner-manager runs the hotel for a salary; the CRE lender underwrites what it would cost to replace them with a third-party operator, because that is the position the lender is in after enforcement.

CRE lenders also frequently substitute the denominator. Rather than the actual 2,680, they size on an underwritten constant — the payment on the loan balance at a stressed rate and a standard amortisation. On a 14,000 balance at 7.50% over 25 years, the annual constant is approximately 1,241, giving an underwritten DSCR of 3,678 / 1,241 = 2.96x. When a broker and a lender quote wildly different DSCRs on the same building, this is usually why.

How does FCCR differ from DSCR?

DSCR asks whether cash covers debt. FCCR asks whether cash covers everything the business is contractually or practically obliged to pay before shareholders see anything.

FCCR = EBITDAR ÷ (Interest + Principal + Rent + Cash tax + Distributions + Unfinanced capex)

FCCR = 5,060 ÷ 5,318 = 0.95x

Below 1.0x. The same borrower that presented at 1.66x does not cover its fixed charges.

Note the structural difference: rent appears on both sides. It is added back into EBITDAR and then charged in the denominator, because a lease is a financing decision dressed as an operating cost. That symmetry is the whole point of the ratio, and it is also why an FCCR is comparable across a borrower who owns its premises and one who leases them, where a DSCR is not.

Drop unfinanced capex out of the denominator — which many US middle-market agreements do — and the same borrower scores 5,060 / 4,698 = 1.08x. One line item, 0.13x. Typical bank FCCR covenants sit at 1.10x–1.25x, so that single drafting choice decides compliance.

Minimum versus average DSCR in project finance

A project finance model produces a DSCR for every period of the debt tenor. Two summary statistics are then covenanted, and they say different things.

• Minimum DSCR = 1.05x (Year 3). This is the binding constraint. It is the year the project is closest to failing, and it is what sizes the debt and the reserve accounts.
• Average DSCR = 1.24x on a ratio-of-sums basis (21,572 / 17,460). On a simple mean of the six annual ratios it is 1.23x. Both are called "average DSCR" and agreements rarely say which. On a lumpier profile the two diverge by considerably more than 0.01x, and the difference has been litigated.

A project sized on average DSCR alone will breach in its worst year by construction. A project sized on minimum DSCR alone is over-collateralised in every other year. Serious sizing uses both, plus a debt service reserve account that carries the shortfall in the trough year, and usually a loan life coverage ratio as a third check.

Pre- versus post-distribution DSCR

The single most consequential distinction on this page, and the one most often left implicit.

Pre-distribution DSCR = 3,142 ÷ 2,680 = 1.17x Post-distribution DSCR = (3,142 − 900) ÷ 2,680 = 2,242 ÷ 2,680 = 0.84x

Northgate paid out 900 to its shareholder. Measured before that payment, it covers debt service comfortably. Measured after, it does not cover it at all.

Which is correct depends on what the ratio is for:

• Testing capacity to pay — pre-distribution. Distributions are discretionary and a lender in difficulty will stop them.
• Testing whether distributions were prudent — post-distribution. This is the basis of a distribution lock-up: the borrower may only pay out if pre-distribution DSCR exceeds a stated level, commonly 1.20x. Northgate at 1.17x fails that lock-up, and the 900 should not have left the company.

An analyst who computes DSCR pre-distribution and never tests the lock-up has missed a covenant breach that is visible in the same spread.

Which DSCR formula should you actually use?

And one rule above all of them: compute the covenant's DSCR from the facility agreement's defined terms, then compute your own analytical DSCR, and show both in the memo. They will not agree. The gap is a credit fact, and it belongs in the recommendation, not in a footnote. The credit assessment memo structure has a place for exactly this.

Where the divergence is being managed across a portfolio rather than one file, the mechanics sit in covenant monitoring in commercial lending, and the full ratio set around DSCR is in the 24 credit analysis ratios that drive a lending decision.

FAQ

How do you calculate the debt service coverage ratio?

Divide the cash available to service debt in a period by the debt service falling due in that period. The general form is settled; what "cash available" and "debt service" mean is not, and that is where every disagreement comes from.

Which DSCR formula should a lender use?

Whichever one the facility agreement defines — that is the only version that can cause a default. Alongside it, compute the version that best reflects how the borrower actually generates cash, and put both in the memo.

Does DSCR include interest or only principal?

Almost always both. A ratio with principal alone is a repayment ratio, not a coverage ratio, and a ratio with interest alone is an interest cover ratio and should be called that.

Why does the same borrower have more than one DSCR?

Because the numerator can be EBITDA, cash flow after tax and capex, net operating income or post-tax cash profit, and the denominator can include or exclude rent, balloons, distributions and personal obligations. Our worked borrower ranges from 0.84x to 1.66x on one set of accounts.

What is a good DSCR?

There is no regulatory answer. Most commercial lenders want 1.20x–1.35x on an EBITDA basis and around 1.15x global for owner-managed credits, but these are policy positions that vary by sector, tenor and security, not standards.

Is FCCR just a stricter DSCR?

Not exactly. FCCR adds rent to both sides of the ratio, which makes it comparable between a borrower who owns its premises and one who leases them. DSCR is not comparable across those two, which is the real reason lease-heavy sectors covenant on FCCR.

Should depreciation be added back in a DSCR?

In an EBITDA or cash-profit DSCR, yes — it is non-cash. But if you add depreciation back and do not then deduct maintenance capex, you have assumed the asset never needs replacing. That is how a 1.66x becomes a 1.17x.

How does IFRS 16 change DSCR?

A borrower that has adopted IFRS 16 reports rent as depreciation and interest instead of an operating cost, so EBITDA rises and the lease liability appears as debt. Two identical businesses on different standards show different DSCRs, which is why many agreements freeze the accounting basis at signing.

What is the difference between minimum and average DSCR?

Minimum DSCR is the worst single period across the debt tenor and is the binding constraint. Average DSCR smooths the profile and always looks better. Agreements that covenant an average without also covenanting a minimum are covenanting almost nothing.

Should distributions be deducted before computing DSCR?

For a capacity test, no. For a distribution lock-up test, yes — that is the entire purpose of the test. Our borrower is 1.17x pre-distribution and 0.84x post, so the answer decides whether a payment was permitted.

Key takeaways

• DSCR is a family of formulas, not one formula. The same borrower scored 0.84x to 1.66x on identical accounts.
• The covenant definition is the only one that can trigger a default. Compute it first, from the defined terms, not from your house template.
• EBITDA DSCR is the most generous variant in the family. Show CFADS beside it or the committee is reading the flattering number.
• The Indian cash-profit convention is a post-tax measure and adds back only term-loan interest — two reasons it prints lower than an EBITDA DSCR on the same accounts.
• FCCR puts rent on both sides, which is the only way to compare a borrower who owns premises with one who leases them.
• Minimum DSCR binds; average DSCR reassures. Covenant both, and say which averaging method you mean.
• Pre- versus post-distribution is the difference between "can it pay?" and "should it have paid out?" A distribution lock-up test needs the second.

Watch YuSight spread a real balance sheet — bring one borrower's accounts and see every DSCR variant computed side by side, each figure cited to its source document and page.

Next: the definitions that cause covenant disputes, covenant monitoring end to end, and the 24 ratios that drive a lending decision.