Why the Waterfall Structure Matters
The NOI waterfall isn't just accounting convention — it's the most direct representation of how a property produces (or fails to produce) income, and the structure of the waterfall reflects the decisions an underwriter has made about risk at each layer. Starting from gross potential rent and working down through vacancy, credit loss, other income, and operating expenses to arrive at NOI forces a level of analytical discipline that a single-line "projected revenue minus projected expenses" summary doesn't.
Every line in the waterfall is an assumption — and assumptions have ranges. Understanding which assumptions are most sensitive to small changes is the analytical work that distinguishes a model that can be stress-tested from one that can only be presented.
Layer 1: Gross Potential Rent
Gross potential rent (GPR) is the income the property would generate if 100% of units or spaces were leased at their scheduled rent for the full period. For stabilized assets, this means the current in-place rent on occupied units plus market rent on vacant units. For value-add assets, this may mean pro forma market rent on some or all units.
GPR calculation requires a clean rent roll: unit count, square footage, in-place rent per unit, and the market rent assumption for vacant or below-market units. The market rent assumption is the first major analytical decision point — use the wrong market rent and every subsequent calculation in the waterfall is wrong.
For multifamily: market rent should be set from actual comparable transactions in the submarket — not asking rents (which are gross of concessions and often stale), not broker opinion, but signed lease comparables from comparable vintage, quality tier, and location. Market rent varies by unit type: 1BR, 2BR, 3BR each have their own market rent range, and a mixed-unit building should use unit-type-specific market rents rather than an average-per-unit figure.
For commercial: market rent PSF annual NNN (or gross, depending on the lease structure) for comparable space in the submarket. The comps should be adjusted for floor, condition, and lease vintage — a ground-floor retail space signed in 2019 is not a clean comp for a second-floor office suite being underwritten today.
Layer 2: Physical Vacancy and Credit Loss
Physical vacancy captures units or spaces that are unoccupied — the lost income from units with no lease in effect. This is where the rent roll vacancy schedule feeds the waterfall: actual vacant units at actual market rent, not a blanket percentage applied to GPR.
Stabilized vacancy assumptions vary by asset class and market. A well-located multifamily asset in a supply-constrained market might underwrite to 4-5% stabilized physical vacancy. A retail center with near-term lease rollover exposure might carry 8-12%. The error pattern to avoid: using a trailing vacancy figure from the rent roll as a forward assumption without adjusting for near-term lease rollover, supply pipeline, or seasonality.
Credit loss is separate from physical vacancy. It captures income from units or spaces that have a lease in effect but where collection is impaired — tenants paying partial rent, in arrears, or in legal proceedings. For stabilized underwriting, credit loss is typically modeled as a small percentage of GPR (0.5-2% depending on asset class and tenant credit profile). For assets with rent-regulated tenants, recovering tenants from pandemic arrears, or commercial tenants with financial stress, a higher credit loss assumption is warranted and should be documented.
One scenario that trips up models: an urban multifamily acquisition with several month-to-month tenants at below-market rents, shown as "occupied" on the rent roll. Physical vacancy is 2%, so GPR to physical vacancy looks clean. But three of those month-to-month tenants have partial-payment histories visible only in the T-12 income data. Credit loss on those units is not 1% — it's running at 15-20% of their scheduled rent. That difference, left in the model as standard credit loss, overstates effective gross income by a meaningful amount.
Layer 3: Other Income
Other income — parking, storage, laundry, pet fees, late fees, application fees, and ancillary revenue streams — is often the line item that gets the least rigorous analysis in acquisition underwriting. It's small relative to base rent, so analysts sometimes use a rounded estimate rather than a line-item build-up from the rent roll and T-12.
For multifamily, other income can represent 5-12% of EGI depending on the property's amenity package and fee structure. Parking income is particularly worth examining: an urban property with structured parking might generate meaningful per-space revenue from both residents and third-party parkers. That income should be modeled at actual current rates and occupancy, not assumed forward at the T-12 average without examining whether the rates reflect current market conditions.
The T-12 other income figures also need to be scrubbed for one-time items. Insurance proceeds, lease buyouts, and other non-recurring items sometimes appear in the "other income" category on an operating statement and need to be stripped out of the forward underwriting. A property showing $45,000 in other income on the T-12 when $20,000 of that was a lease termination payment has a sustainable other income base of $25,000.
Layer 4: Effective Gross Income and the Concession Adjustment
Effective gross income (EGI) is GPR minus vacancy and credit loss, plus other income. It represents the income the property is actually expected to collect. In environments with significant lease concessions — free rent, move-in credits, or lease-up specials — EGI should reflect the net collected income, not the gross scheduled rent.
The concession adjustment is the place where acquisition models built from rent rolls (which show scheduled rent) diverge from models built from actual cash flow data. A multifamily property in a lease-up phase with 15% of units on concession packages averaging 6 weeks free rent is not generating EGI equal to 100% of scheduled rent. The waterfall should reflect the actual concession burn in the current period and forward assumptions on concession levels as the market evolves.
Layer 5: Operating Expenses and the Expense Lookback Problem
Operating expenses — property taxes, insurance, management fees, maintenance and repairs, utilities, administrative costs — should be underwritten from T-12 actuals, adjusted for known non-recurring items and known forward changes (property tax reassessment on a post-sale basis is the most important forward adjustment).
T-12 expense lookback discrepancies are common and consequential. Three patterns that appear regularly in diligence: (1) deferred maintenance — the T-12 shows low repair and maintenance expense because the current owner has been deferring work; the forward maintenance budget needs to reflect catch-up CapEx and a normalized ongoing level. (2) Insurance premium reset — insurance costs have increased significantly in many markets; a T-12 from 24 months ago may show insurance costs 30-50% below what the buyer will actually pay. (3) Management fee understating — the current owner self-manages and applies a minimal management fee; the buyer, who will use a third-party manager, needs to use market management fee rates (3-6% of EGI for multifamily depending on asset size).
The management fee issue is particularly common in smaller multifamily acquisitions where the seller is an individual owner-operator. Plugging in a management fee that matches a professional property management rate rather than the self-management cost can reduce NOI by enough to meaningfully alter the going-in cap rate. We're not saying the seller's self-management approach is illegitimate — it's their prerogative. We're saying the buyer's lender will normalize to market management fees regardless, and the underwriting should too.
Layer 6: Arriving at NOI and What It Tells You
NOI = Effective Gross Income minus Operating Expenses. It excludes debt service, depreciation, capital expenditures, and income taxes — it's the pre-leverage income produced by the asset on a recurring cash flow basis.
The NOI figure that comes out of a properly built waterfall is the figure that gets capitalized to determine value, used to size debt, and stress-tested in the sensitivity analysis. Getting it right means building each layer from defensible inputs, documenting assumptions explicitly, and reconciling against T-12 actuals at each line.
The sensitivity table that belongs attached to every NOI waterfall should show, at minimum: NOI at base case, at +2% and -5% on revenue, at +5% and +10% on expenses, and at combinations of stress scenarios. Running a single-point NOI estimate into a DSCR and cap rate analysis without a sensitivity envelope is presenting false precision to the IC.
Building the NOI waterfall right takes longer than plugging numbers into a template. The acquisition team that does it correctly understands the asset's income mechanics at a level that becomes valuable in negotiation, in financing conversations, and in the asset management decisions made after closing. That understanding lives in the structure of the waterfall — which is why the methodology matters as much as the final number.
