Claude AI for Commercial Loan Underwriting: Credit Memos, DSCR, and Covenant Analysis

Claude AI for Commercial Bankers and Credit Analysts

Commercial loan underwriting is the backbone of bank lending — but it's also one of the most time-consuming workflows in finance. Writing a thorough credit memo, spreading three years of borrower financials, calculating DSCR across scenarios, and structuring a covenant package used to take days. With ClaudeFinLab's Commercial Banking SKILL.md templates, Claude produces loan-ready analysis in minutes.

AI adoption in commercial banking is accelerating fast — banks using AI for credit analysis report 50–75% reduction in underwriting cycle time. Anthropic's Claude has emerged as the preferred tool for credit professionals because of its ability to maintain complex financial logic across a long conversation and produce consistently structured, professional output. ClaudeFinLab provides the structured prompt frameworks (SKILL.md templates) that make Claude behave like a trained credit officer, not just a general-purpose assistant.

Writing Commercial Credit Memos with Claude AI

The credit memo is the central deliverable in commercial loan underwriting. A well-written memo covers borrower background, financial analysis, risk assessment, collateral, and a clear recommendation. Claude can produce a complete, structured credit memo from raw financial inputs.

Sample Prompts for Credit Memo Writing

• "Write a complete commercial credit memo for a $3.5M 7-year term loan to a regional HVAC services company. Revenue $9.2M, EBITDA $1.38M, net income $820K, D&A $180K. Collateral: commercial real estate (appraised $4.1M, existing mortgage $1.6M), business assets. Personal guarantee from 100% owner (net worth $3.2M). Cover: executive summary, borrower profile, financial analysis (include 3-year ratios), DSCR, risk factors, collateral, and recommendation."
• "Draft a credit memo for a $750K operating line of credit for a medical practice. 3-year average revenue $4.2M, average EBITDA $620K. Discuss key risks for medical practice lending: payer mix concentration, regulatory changes, key-person risk. Structure the memo with risk rating: Pass, Special Mention, Substandard."
• "I'm presenting to the credit committee tomorrow. Write a one-page executive summary credit memo for a $12M leveraged buyout loan to acquire a distribution business. EV/EBITDA: 5.2x. Sponsor contribution: 35% equity. Projected DSCR: 1.42x base case, 1.18x stress. Three key strengths and three key risks."

DSCR Analysis and Debt Structuring with Claude AI

Debt Service Coverage Ratio (DSCR) is the primary cash flow metric for commercial lending. Claude AI can calculate DSCR, build sensitivity tables across rate and NOI scenarios, and recommend structuring adjustments to meet minimum coverage thresholds.

• "DSCR calculation: Commercial real estate property, NOI $318K. Proposed loan $3.2M, 25-year amortization, 6.875% rate. Annual debt service = $3.2M × 7.89% mortgage constant = $252K. DSCR = $318K / $252K = 1.26x. Bank minimum 1.25x — barely passes. Run a 5×5 sensitivity table: rows = NOI from $280K to $360K in $20K steps; columns = interest rate from 6.0% to 8.0% in 50bps steps. Flag all cells where DSCR falls below 1.25x."
• "Debt yield and LTV underwriting for an office building: NOI $485K, purchase price $6.8M at a 7.1% going-in cap rate. Proposed loan $4.5M. Debt Yield = $485K / $4.5M = 10.8%. LTV = $4.5M / $6.8M = 66.2%. Run through the lender underwriting box: DSCR must be ≥ 1.25x, Debt Yield ≥ 9.0%, LTV ≤ 70%. Does this loan pass? What's the maximum loan at each constraint?"
• "Global cash flow underwriting for an S-Corp owner: Business EBITDA $1.1M, owner's salary $220K (already in EBITDA). Proposed new term loan: $480K/year in debt service. Personal obligations: primary residence mortgage $54K/year, car loans $28K/year. Calculate: (1) business DSCR net of all business debt, (2) global DSCR including personal obligations, (3) federal and state tax estimate to subtract from cash flow, (4) whether global DSC meets 1.20x minimum."

Borrower Financial Spreading and Ratio Analysis

Spreading financial statements — converting raw borrower financials into standardized credit ratios — is one of the most labor-intensive tasks in commercial lending. Claude AI normalizes and recast financials, identifying owner adjustments and non-recurring items automatically.

• "Spread these 3 years of financials for a $4M business acquisition loan. I'll paste the income statement and balance sheet for FY2023, FY2024, FY2025. Produce a standardized spreading table with: revenue, gross profit, EBITDA (computed and adjusted), net income, total assets, total debt, equity. Compute: leverage (Debt/EBITDA), DSCR, current ratio, gross margin, EBITDA margin, revenue CAGR. Recast EBITDA by adding back: $85K one-time legal settlement FY2024, $45K excess owner compensation above market."
• "Annual ratio trend analysis for a 5-year credit review. Compute and present in a table: (1) profitability — gross margin, EBITDA margin, ROA, ROE; (2) leverage — Debt/EBITDA, Debt/Equity; (3) liquidity — current ratio, quick ratio, cash ratio; (4) coverage — interest coverage (EBIT/interest), DSCR. Provide a 3-sentence trend commentary for each category."

Covenant Structuring and Compliance Tracking

Financial covenants protect the lender throughout the loan term. Claude AI helps bankers design covenant packages and test borrower compliance at each reporting period.

• "Design a covenant package for a $6M revolving credit facility to a commercial printing company (cyclical industry, EBITDA $1.1M, existing debt $2.8M). Include: (1) DSCR covenant tested quarterly with the right definition, (2) maximum leverage, (3) minimum liquidity, (4) CapEx limit. Provide the full covenant definition language and the current company's headroom against each covenant."
• "Covenant compliance certificate: As of 9/30/26 — DSCR covenant ≥ 1.25x: computed DSCR 1.21x (BREACH). Leverage ≤ 4.0x: actual 3.6x (pass). Minimum liquidity $500K: actual $820K (pass). Draft a bank notification memo documenting the DSCR breach, analysis of whether it's a technical or material breach, the cure period, and a proposed waiver request with management commentary."

SBA Loan Analysis and Narrative

The Small Business Administration (SBA) 7(a) and 504 programs serve millions of small business borrowers annually. Claude AI helps SBA lenders assess eligibility, structure the loan, and write the credit narrative.

• "SBA 7(a) loan for a restaurant acquisition: Purchase price $1.1M. Seller's discretionary earnings (SDE) $285K (3-year average). Proposed SBA loan: $900K, 10-year term, current prime-based rate (approximately 11.25%). Annual debt service $147K. DSCR = $285K / $147K = 1.94x. Equity injection: $200K from buyer (18.2%). Assess: (1) SBA eligibility — is this a small business? (2) Is the equity injection adequate? (3) DSCR is strong but is SDE the right cash flow measure? Discuss risks: restaurant industry, working capital, key-person. Recommend."
• "SBA 504 project for owner-occupied commercial real estate: Total project cost $2.4M (building $1.9M + equipment $500K). SBA 504 structure: CDC debenture 40% ($960K at fixed rate), bank first mortgage 50% ($1.2M), borrower equity 10% ($240K). Calculate: bank DSCR using only first mortgage payments; combined DSCR on total debt service. Is the deal bankable? Write the SBA-required credit narrative."

How to Use ClaudeFinLab's Commercial Banking Skills

ClaudeFinLab's Commercial Banking category includes seven SKILL.md templates covering credit memos, covenant compliance, financial spreading, DSCR modeling, call prep briefs, and SBA loan narratives. Each template loads into Claude as a project system prompt — once loaded, Claude follows professional banking workflows automatically for any borrower or loan scenario you describe.

To get started: browse the Commercial Banking templates at /category/commercial-banking/, copy any template into Claude's Projects system prompt, and describe your borrower. Claude will follow the credit analysis workflow step by step.

The Future of AI in Commercial Lending

Banks that have piloted AI-assisted underwriting report dramatic efficiency gains: credit memo preparation time reduced from 4–8 hours to under 1 hour; DSCR modeling that previously required dedicated financial models now handled in a conversational prompt. The next evolution is fully agentic credit workflows — where Claude reads loan documents, pulls financial data from bank systems, performs analysis, and routes applications — but structured SKILL.md prompt templates are the practical starting point available today.