Underwrite Asset

Underwrite Asset

Build an institutional-grade underwriting package for an income-producing property.

Skills Invoked

• underwriting-standards -- pro forma construction, NOI projection, cash flow waterfall
• rent-roll-normalization -- rent roll cleanup, rollover, WALT, mark-to-market
• valuation-methods -- cap rate and DCF valuation with supportable range
• market-analysis -- submarket context for rent/vacancy assumptions
• debt-sizing-hierarchy (capital-markets plugin) -- loan sizing if debt assumptions provided

Required Inputs

• Rent roll (Excel or PDF) OR stated rent/occupancy assumptions
• Trailing 12-month operating statement (T-12) OR stated expense assumptions

Optional Inputs

• Offering memorandum (PDF) -- will extract property details and seller assumptions
• Market data -- submarket, property type, class
• Debt terms -- loan amount, rate, term, amortization
• Hold period and exit cap rate assumptions

Reads from Manifest

None -- this is typically the first command in an acquisition workflow.

Writes to Manifest

• Directory: underwriting/
• Files: proforma.xlsx, assumptions.json, summary.md
• Key metrics: going_in_cap, stabilized_noi, unlevered_irr, levered_irr, price_per_sf, price_per_unit

Workflow

Step 1: Gather Property Information

If the user provided an OM, extract: property name, address, type, size (SF/units), year built, seller's asking price. If not, ask the user for these details.

Step 2: Normalize the Rent Roll

Invoke rent-roll-normalization to clean and standardize the rent roll. Output: in-place rents, vacancy, rollover schedule, WALT, top-tenant concentration.

Step 3: Analyze Trailing Operations

If a T-12 is provided, parse revenue and expense line items. Compare in-place revenue to rent roll. Flag discrepancies.

Step 4: Build the Pro Forma

Invoke underwriting-standards to construct a multi-year pro forma. The skill will call validated Python functions in scripts/proforma.py for all calculations. Key outputs: revenue projection, expense projection, NOI, capital reserves, cash flow before/after debt.

CRITICAL: Do NOT calculate NOI, IRR, or cash flows by reasoning through the math. Invoke the Python functions.

Step 5: Size Debt (if applicable)

If debt assumptions are provided, invoke debt-sizing-hierarchy to test DSCR, LTV, and debt yield constraints.

Step 6: Calculate Returns

Invoke underwriting-standards which calls scripts/returns.py for IRR, equity multiple, and cash-on-cash yield. Do NOT compute these manually.

Step 7: Value the Asset

Invoke valuation-methods to produce a cap rate valuation and DCF valuation. Cross-check against per-SF/per-unit benchmarks.

Step 8: Produce Output

Generate proforma.xlsx using the template in underwriting-standards/assets/proforma-template.xlsx. Fill named ranges with computed values. Include "Sources & Assumptions" tab.

Generate assumptions.json with all inputs and assumptions used.

Generate summary.md with executive summary of the underwriting.

Write all to the deal context directory and update _manifest.json.

QA Checklist

• Rent roll total ties to T-12 revenue (within 2%)
• Expense growth assumptions stated and sourced
• Cap rate supported by market comps or stated as assumption
• Vacancy assumption justified (physical vs. economic)
• Capital reserves included
• IRR and equity multiple computed by validated code, not LLM reasoning
• Sources & Assumptions tab populated

Escalation Notes

• If rent roll is missing or incomplete, ask the user before proceeding
• If T-12 is unavailable, proceed with rent roll only but flag in summary
• If market data is unavailable, note the gap and use manual assumptions
• If rent roll and T-12 revenue diverge by >5%, flag and ask user which to rely on