DSCR Underwriting: How Lenders Actually Calculate It
How middle-market underwriters compute DSCR on $1M–$10M raises — the three formulas that disagree, the pro-forma debt stack they build, and four reconciliations that keep files inside covenant.
The number every underwriter opens with
Before a credit officer reads your narrative, they run one number: Debt Service Coverage Ratio (DSCR). It is the single most weighted line in most middle-market credit memos, and it is where the majority of "declined at committee" files break. Yet borrowers routinely present a DSCR their lender does not recognize — computed off the wrong EBITDA, missing the wrong add-backs, or ignoring the wrong pro-forma debt.
This briefing walks through how underwriters actually calculate DSCR on a $1M–$10M raise, where borrower math diverges from lender math, and the four defensible reconciliations that keep a file inside covenant on day one.
The formula, and where it splits
The textbook definition is clean:
DSCR = Cash Flow Available for Debt Service (CFADS) ÷ Total Debt Service
The disagreement is entirely in the numerator. Three definitions dominate:
- Bank DSCR (regulated lenders, SBA 7(a)): EBITDA − unfinanced CapEx − cash taxes − distributions to owners.
- Private credit DSCR: Adjusted EBITDA − maintenance CapEx (contractually defined) − cash taxes.
- Borrower DSCR (what most CIMs show): EBITDA ÷ interest + principal, with no CapEx and no tax burden netted out.
The third one is not wrong — it's simply not what committee will use. A file that leads with a 1.8× "borrower DSCR" but pencils to a 1.05× "bank DSCR" is not a strong file. It is a file that will be re-priced, structurally covenanted, or declined outright.
The denominator — what counts as debt service
Total Debt Service is not just the new facility. Underwriters build a pro-forma debt schedule that stacks:
- Interest and principal amortization on the proposed facility (usually stressed +100–200 bps above the indicative rate)
- Existing term debt not being refinanced
- Capitalized lease obligations (ASC 842 operating leases in most bank models)
- Contingent obligations that "look like debt" — earnouts with fixed floors, deferred purchase price, seller notes
- Owner-guaranteed obligations on affiliate entities if consolidation is warranted
If your DSCR model excludes any of these, expect it to fail underwriting recalculation. The most common miss is capitalized leases. A logistics business with $400K/year in truck leases can lose 25–40 bps of DSCR when the lender adds them back into debt service.
The four defensible reconciliations
The strongest files don't fight the lender's math — they pre-reconcile to it. Four schedules do most of the work:
1. EBITDA-to-CFADS bridge
Start at reported EBITDA. Show, line by line, the deductions to arrive at cash flow available for debt service: unfinanced CapEx (three-year trailing average is the standard rebuttal to a spike year), cash taxes (federal + state, at the entity's actual effective rate, not statutory), and any mandatory distributions. Do NOT net out discretionary owner comp — that belongs in the add-back schedule, not the CFADS deduction.
2. Debt service stress table
Show DSCR at the indicative rate, +100 bps, and +200 bps. Committee will run this themselves — better to run it for them, at your model, with your assumptions locked in. A file that penciles to 1.35× at +200 bps of stress is fundamentally different from one that penciles to 0.95×, and pretending otherwise wastes underwriting time.
3. Covenant headroom analysis
The proposed facility will carry a DSCR covenant, typically 1.20× or 1.25× for lower-middle-market senior debt. Compute your headroom — projected DSCR minus covenant DSCR — for each of the next 8 quarters. Highlight the tightest quarter. If headroom drops below 15 bps in any quarter, the file needs restructuring before it goes to committee, not after.
4. Maintenance vs. growth CapEx split
Private credit facilities almost always define "maintenance CapEx" contractually. Bank facilities usually let unfinanced CapEx flow through unadjusted. The distinction matters: a manufacturer investing $1.2M in a new line that expands capacity should be able to defend $400K as maintenance and $800K as growth, with vendor invoices and a written maintenance schedule as backup.
Three failure modes that show up in committee
The single-year EBITDA claim. A trailing-twelve-month EBITDA that includes one anomalous quarter (a large one-time contract, a delayed expense) is not what committee will underwrite. Expect them to normalize to a three-year weighted average, or to LTM ex-anomaly. Present both proactively.
The forgotten seller note. In acquisition financings, borrowers frequently exclude the seller note from debt service, treating it as equity-like. Some lenders will accept subordination and exclude it; most will haircut it 50% or include it fully. Ask early, structure accordingly.
The uncoded distribution. S-corp and LLC borrowers sometimes book owner tax distributions as "operating expense" rather than distributions. That inflates EBITDA and understates the CFADS deduction simultaneously. Underwriters catch this in the tax return reconciliation, and it damages credibility for the rest of the file.
What "bankable DSCR" actually means
A bankable DSCR is not a number — it is a defensible reconciliation. A 1.28× DSCR that ties cleanly to tax returns, three years of audited financials, a documented maintenance CapEx policy, and a stressed debt service table will clear committee. A 1.75× DSCR built on adjustments the borrower cannot substantiate will not.
The scan below runs the same reconciliation an underwriter will run on your file — before you send it out.
See what a lender sees — before you ever apply.
Run the free scan. Get your Capital Readiness Score and Gap Report in about a minute.
SBA 7(a) vs. Private Credit: Which Fits a $1M–$10M Raise
The committee-readiness playbook for $1M–$10M raises. SBA 7(a) is a compliance underwrite; lower-middle-market private credit is a growth underwrite. Which lane a business belongs in, and how to build a file that clears the committee it actually sits in front of.
The CIM That Closes: What Credit Committee Actually Reads
Committee is paid to answer four questions. A committee-ready CIM answers all four inside the first five pages — and self-identifies three real risks with three real mitigants.
Adjusted EBITDA: What Underwriters Add Back — And What They Don't
The most negotiated line in every credit file. Which add-backs survive quality-of-earnings review, which get rejected, and how committee actually reads the reconciliation.
Institutional capital advisors — 60+ combined years across SBA, private credit, and M&A.