Guide
Commercial Real Estate Underwriting: A Complete Guide
Commercial real estate underwriting is the analytical process of estimating a property's future cash flows and then testing whether those cash flows can service debt and clear an investor's required return at an acceptable level of risk. In practice it means building a defensible pro forma — stabilized income minus operating expenses equals net operating income (NOI), which drives value, debt capacity, and levered equity returns — and stress-testing every assumption behind it. Done well, underwriting converts a broker's offering memorandum into a set of numbers you can defend to an investment committee.
Start with income: gross potential rent to effective gross income
Underwriting is built from the top line down. Begin with gross potential rent (GPR) — every unit or square foot leased at market rent, whether occupied or not — then subtract loss-to-lease, vacancy, credit loss, and concessions to reach effective gross rental income. Add other income (parking, RUBS/utility reimbursements, storage, laundry, late fees, percentage rent in retail) to get effective gross income (EGI). The single most common way to over-pay is to underwrite in-place rents up to an aspirational market rent without a defensible comp set and a realistic lease-up schedule. Verify every current rent against the rent roll, and validate market rent against leased comps you can name, not asking rents.
- •GPR = all units × market rent (100% occupancy assumption)
- •Less: physical vacancy, credit/collection loss, loss-to-lease, concessions
- •Plus: other income (reimbursements, parking, fees, ancillary revenue)
- •= Effective Gross Income (EGI) — the real revenue the asset collects
- •Reconcile every line to the actual rent roll and trailing income statement, not the broker's stabilized pro forma
Operating expenses and NOI: the number that sets value
From EGI, subtract operating expenses to reach net operating income (NOI). Operating expenses include property taxes (re-assess at your purchase price, not the seller's basis — a frequent and costly miss), insurance, utilities, repairs and maintenance, payroll, management fees, and administrative/marketing costs, plus a non-financeable replacement reserve. NOI explicitly excludes debt service, capital expenditures, depreciation, and income taxes — keep those below the line so NOI stays capital-structure-neutral and comparable across deals. NOI is the fulcrum of the entire model: it sets value via the cap rate (Value = NOI ÷ cap rate), and it drives every debt-sizing test. A 5% error in expenses can move value by multiples of that at a low cap rate, so ground each line in the trailing-12 (T-12) operating statement and adjust only where you can justify the change.
- •NOI = EGI − operating expenses (before debt service and capital items)
- •Re-underwrite property taxes to your acquisition basis where reassessment applies
- •Include a replacement reserve; exclude one-time capex, tenant improvements, and leasing commissions from NOI
- •Sanity-check your expense ratio and per-unit/per-SF expenses against comparable operating assets
Debt sizing: how much leverage the deal actually supports
Lenders don't lend to your target return — they lend to the smallest loan amount that passes their constraints, and you should size debt the same way. Run all four standard tests and take the minimum: loan-to-value (LTV, loan ÷ value), loan-to-cost (LTC, loan ÷ total project cost, which governs on development and heavy value-add), debt service coverage ratio (DSCR, NOI ÷ annual debt service, typically a 1.20x–1.35x floor), and debt yield (NOI ÷ loan amount, a value-independent leverage cap lenders lean on hardest when cap rates are compressed). Model the actual debt terms — interest-only period, amortization, index plus spread, and any rate cap on floating debt — because the constraint that binds shifts with interest rates. Advanced structures (preferred equity, mezzanine/junior debt, and a mid-hold refinance to return capital) sit on top of this senior-debt analysis and change the equity story materially.
- •Size to the MINIMUM loan across LTV, LTC, DSCR, and debt yield
- •DSCR = NOI ÷ annual debt service; debt yield = NOI ÷ loan — debt yield ignores cap rate and is the honest leverage governor
- •Model IO periods, amortization, and floating-rate caps explicitly — the binding constraint moves with rates
- •Layer preferred equity or mezz debt only after the senior loan is sized
Equity returns and the waterfall: translating cash flow into investor outcomes
Once debt is sized, the levered cash flow after debt service is what equity actually receives. Evaluate it on the core return metrics — internal rate of return (IRR, money-weighted because it accounts for the timing and size of every cash flow, and the primary hurdle for most sponsors), equity multiple (total cash returned ÷ invested, a magnitude check IRR alone hides), cash-on-cash return (annual pre-tax cash flow ÷ invested equity, the year-by-year yield), and yield-on-cost (stabilized NOI ÷ total cost, compared against the exit cap rate to measure development/value-add spread). In a GP/LP structure, those returns are then split through an equity waterfall: limited partners typically receive a preferred return first, capital is returned, a GP catch-up may follow, and residual profit is split via a promote (carried interest) at rising IRR hurdle tiers — often with a clawback protecting LPs at the end. Model the waterfall precisely; the difference between a 7% and an 8% pref, or whether the promote is on 70/30 above an 8% or a 12% hurdle, can swing LP IRR by hundreds of basis points.
- •IRR (money-weighted) and equity multiple (total return) — always report both
- •Cash-on-cash = annual pre-tax cash flow ÷ equity invested
- •Waterfall order: preferred return → return of capital → GP catch-up → tiered promote split, with clawback backstop
- •Distinguish LP-level returns from deal-level (unlevered) returns when presenting to investors
Risk, scenarios, and sensitivity: pressure-testing the base case
A single base-case pro forma is a forecast, not underwriting. The analytical work is in the downside: build bear, base, and bull scenarios, and run a two-variable sensitivity table on the assumptions that actually move returns — most often exit cap rate against rent growth or purchase price, and interest rate against DSCR. Because value is NOI ÷ cap rate, exit cap rate is usually the single largest driver of returns and the one you control least, so underwrite exit cap expansion (a higher exit cap than entry) as your default conservative posture. Test break-evens: the vacancy or expense level at which DSCR breaches 1.0x, the rate at which the loan no longer sizes, the rent growth required to hit your target IRR. The deals that fail are rarely the ones with a bad base case — they're the ones whose base case only works if three optimistic assumptions all come true at once.
- •Model bear / base / bull scenarios, not just a single line
- •Two-variable sensitivity: exit cap × rent growth, and interest rate × DSCR, are the highest-value tables
- •Underwrite exit cap ≥ entry cap as a default conservative stance
- •Find break-even occupancy, break-even rate, and the DSCR-1.0x cliff before you commit capital
Asset-class nuance and documenting your work
The NOI-to-returns skeleton is universal, but the income side differs sharply by property type, and using the wrong template produces confident, wrong numbers. Multifamily and self-storage underwrite on short-term leases and unit/occupancy dynamics; retail, office, and industrial run on a rent roll of longer-term leases with tenant credit, lease expiration schedules, TI/LC costs, and reimbursement structures (NNN, modified gross, full-service gross); hospitality underwrites on RevPAR, ADR, and department-level P&L rather than leases. Whatever the asset class, the deliverable is an investment committee memo that ties every headline number back to a source — the rent roll, the T-12, the comp set, the loan term sheet. That traceability is the difference between an underwriting model and a spreadsheet of opinions. Origentic runs all seven asset classes on one deterministic calc engine — the math lives in code and is verified against a frozen test-vector suite, so every number carries a source, and AI document extraction pulls rent rolls and T-12s with per-field provenance and mandatory human sign-off before anything reaches your model.
- •Lease-based assets (retail/office/industrial): model rollover, TI/LC, and reimbursement type per tenant
- •Occupancy-based assets (multifamily/self-storage): focus on lease-up, turnover, and market-rent comps
- •Hospitality: underwrite on RevPAR/ADR and departmental margins, not leases
- •Ship an IC memo where every figure traces to a source document, not a rounded assumption
Frequently asked questions
- What is the difference between NOI and cash flow?
- NOI (net operating income) is effective gross income minus operating expenses, before any financing or capital items. Cash flow — specifically levered cash flow — is what remains after subtracting debt service and capital expenditures from NOI. NOI is deliberately capital-structure-neutral so you can compare assets and derive value (Value = NOI ÷ cap rate); cash flow is what equity investors actually receive and is what drives IRR, equity multiple, and cash-on-cash return.
- How do lenders decide the maximum loan amount?
- They run several tests and lend the smallest amount that satisfies all of them: loan-to-value (LTV), loan-to-cost (LTC), a minimum debt service coverage ratio (DSCR, usually 1.20x–1.35x), and a minimum debt yield (NOI ÷ loan). When values are high and cap rates low, debt yield or DSCR typically binds before LTV. Always size to the minimum across all four rather than assuming your target leverage is available.
- Why is the exit cap rate so important in underwriting?
- Because reversion value equals stabilized NOI divided by the exit cap rate, small changes in the exit cap move sale proceeds — and therefore IRR and equity multiple — more than almost any other assumption. It's also the variable you control least, since it depends on market conditions years out. Conservative underwriting assumes the exit cap is equal to or higher than the entry cap, and stress-tests exit cap expansion in a two-variable sensitivity table.
- What makes commercial underwriting different across asset classes?
- The return framework (NOI → value → debt → equity returns) is the same, but the income side differs. Multifamily and self-storage run on short leases and occupancy; retail, office, and industrial run on tenant-level rent rolls with lease rollover, TI/LC costs, and reimbursement structures; hospitality underwrites on RevPAR and ADR rather than leases. Using the correct income model for each type is essential — one engine handling all seven asset classes keeps the returns math consistent while respecting these differences.
Keep exploring