




Every multifamily deal starts with a number someone put into a spreadsheet. How that number was arrived at, and how honestly the assumptions behind it were stress tested, is what separates deals that perform from deals that disappoint.
Good underwriting doesn't guarantee a good outcome. Markets move, tenants leave, expenses surprise you. But disciplined underwriting significantly improves your odds by forcing you to think clearly about risk before you commit capital, rather than after. It's the foundation every sound multifamily investment decision is built on, and it's worth understanding whether you're buying your first duplex or underwriting a 300-unit acquisition for an institutional fund.
This guide covers everything you need to know about multifamily underwriting in 2026, from the core metrics and assumptions to the step-by-step process, the mistakes that most commonly derail deals, and how lender underwriting differs from investor underwriting.
Related:
Multifamily underwriting is the process of analyzing an apartment property's income, expenses, financing, and projected returns to determine whether it represents a sound investment. It's the work that happens before a deal closes, when an investor or lender builds a financial model to answer the fundamental questions: what is this property worth, what can it support in debt, and what will it return to equity investors over the hold period?
At its core, underwriting is about converting a set of assumptions into numbers, and then stress testing those numbers against different scenarios to understand how the deal performs if things don't go exactly to plan. Every line item in a multifamily underwriting model, from current rents to projected vacancy to exit cap rate, represents a judgment call about the future. The quality of those judgment calls determines whether the model reflects reality or just tells you what you want to hear.
Underwriting happens at two levels in any multifamily transaction. Equity investors underwrite to determine whether the deal will generate their target returns, typically measured by cash-on-cash return, IRR, and equity multiple. Lenders underwrite to determine whether the property generates enough income to support the requested loan amount, measured primarily by DSCR, LTV, and debt yield. Both processes use much of the same underlying data, but they ask different questions and reach different conclusions.
Gross potential rent is the total income a property would generate if every unit were occupied and paying market rent for the full year. It's the starting point for every multifamily underwriting model and the top line number everything else flows from. GPR is calculated by multiplying each unit type's market rent by the number of units and by 12 months.
Vacancy and credit loss is the deduction applied to GPR to account for units that are unoccupied and rent that goes uncollected due to non-payment or bad debt. Most underwriters apply a combined vacancy and credit loss rate of 5% to 10% of GPR depending on the property type, market, and current occupancy trends. The result is effective gross income, the revenue the property can realistically be expected to collect.
Net operating income is the single most important number in multifamily underwriting. It's what's left after you subtract all operating expenses from effective gross income, and it's the foundation for almost every other metric in the model including valuation, debt sizing, and return calculations. A property's NOI drives its value through the cap rate formula, determines how much debt it can support through the DSCR calculation, and anchors the cash flow projections that feed into IRR and equity multiple.
The capitalization rate, or cap rate, is the ratio of a property's NOI to its market value or purchase price. It's used both to value a property at acquisition and to project its value at exit, making it one of the most consequential assumptions in any multifamily underwriting model. A property generating $500,000 in NOI valued at a 5.0% cap rate is worth $10 million. The same NOI at a 5.5% cap rate is worth roughly $9.1 million, a $900,000 difference from a single half-point move in the cap rate.
Debt service coverage ratio measures the property's ability to cover its loan payments from operating income. It's calculated by dividing NOI by total annual debt service, and it's the primary metric lenders use to determine how much they're willing to lend on a multifamily property. Most agency lenders require a minimum DSCR of 1.25, meaning the property must generate 25% more income than its debt payments require.
Cash-on-cash return measures the annual pre-tax cash flow a property generates as a percentage of the total equity invested. It's calculated by dividing annual cash flow after debt service by total equity invested, and it tells investors what they're earning on their capital each year from operations alone, excluding any proceeds from a future sale or refinance.
Read Also:
Multifamily underwriting follows a logical sequence. Each step builds on the one before it, and errors made early in the process compound through everything that follows. Here's how to work through a deal from the first number to the final return calculation.
Before you build a single formula in your model, get your hands on the property's actual financial data and verify it. This means requesting the current rent roll, the trailing 12-month operating statements, the current lease agreements, and any available information on recent capital expenditures and deferred maintenance.
The rent roll tells you who is in each unit, what they're paying, when their lease expires, and how current or below-market their rent is relative to what comparable units are achieving. The operating statements tell you what the property has actually been spending on taxes, insurance, utilities, maintenance, and management. Neither document should be taken at face value without verification.
Cross-check the rent roll against the lease agreements, compare the reported expenses against market benchmarks for similar properties, and flag anything that looks unusually low or inconsistently reported.
With the rent roll in hand, build your revenue model starting with gross potential rent. Calculate GPR by multiplying each unit type's market rent by the number of units and by 12 months. Use actual comparable lease data from the submarket to anchor your market rent assumptions, not the broker's opinion of what rents could be with the right management.
If the property has below-market leases, model the burn-off carefully. Tenants on leases below market rent won't immediately start paying market rents when their leases expire. Some will renew at a negotiated rate below market. Some will leave and need to be replaced, which triggers a vacancy period and a turnover cost. Modeling this transition realistically rather than assuming an immediate jump to market rents is one of the marks of disciplined underwriting.
Operating expenses are where a lot of underwriting models get optimistic in ways that aren't immediately obvious. The most common error is underestimating expenses by using the seller's reported numbers without adjusting for items that are missing, understated, or one-time in nature.
Work through each expense category line by line. Property taxes deserve particular attention because they often reset at a higher level after a sale in many jurisdictions, meaning the seller's current tax bill may be significantly lower than what you'll pay after closing. Insurance costs have risen sharply in many markets in recent years and should be verified with an actual quote rather than extrapolated from the seller's current premium. Management fees should reflect what a professional third-party manager would charge, even if you plan to self-manage, because the fee represents the true economic cost of that function.
Build in a replacement reserve line for ongoing capital expenditures even if the property is in good condition. Most institutional underwriters use $200 to $400 per unit per year as a baseline replacement reserve, with higher amounts for older properties or those with deferred maintenance. Leaving this line out of the model doesn't make the expense disappear. It just means you'll be surprised by it later.
Subtract your total operating expenses from effective gross income to arrive at NOI. This is the number that drives everything else in the model, so spend time making sure it's right before you move forward.
Divide NOI by the market cap rate for comparable properties in the submarket to arrive at your estimate of the property's market value. This is your primary valuation check against the purchase price. If the seller is asking $8 million for a property with $400,000 in NOI and comparable properties are trading at 5.0% cap rates, the implied value is $8 million and the deal is priced at market. If comparable cap rates are 5.5%, the implied value is $7.27 million and the seller is asking for a premium that needs to be justified by something, usually a value-add business plan that will push NOI higher.
With your NOI established, model the debt structure. Start with the loan amount, which will be constrained by whichever is more binding: the LTV limit or the DSCR minimum. Run both calculations and use the lower of the two loan amounts.
For the LTV calculation, multiply the purchase price or appraised value by the lender's maximum LTV ratio, typically 65% to 80% depending on the loan type. For the DSCR calculation, determine the maximum annual debt service the NOI can support at the lender's minimum coverage ratio, then back into the loan amount that produces that debt service at your assumed interest rate and amortization period.
Model the full debt service schedule over the hold period, including any rate adjustments if you're using floating rate debt. Rising rates on a floating rate bridge loan can meaningfully affect cash flow projections in years two and three of the hold, and that impact needs to be visible in the model before you close.
With revenue, expenses, NOI, and debt service all modeled, you can calculate the returns the deal generates for equity investors. The key metrics to calculate are cash-on-cash return for each year of the hold period, total equity multiple across the full hold, and IRR incorporating the timing of all cash flows in and out of the deal.
Project the property's value at exit by applying your assumed exit cap rate to the projected NOI in the year of sale. Subtract the outstanding loan balance, closing costs, and any disposition fees to arrive at net sale proceeds. Add those proceeds to the cumulative cash distributions paid during the hold period to get total distributions, then divide by total equity invested to calculate the equity multiple.
Run the full IRR calculation incorporating the initial equity investment as a negative cash flow, each year's distributable cash flow as a positive cash flow, and the net sale proceeds in the exit year as the final positive cash flow. The resulting IRR is your annualized return on invested capital across the full hold period.
A single set of base case assumptions is not enough. Before you make a decision based on your underwriting model, run it through at least two additional scenarios: a downside case and a severe downside case.
The downside case might assume rents come in 5% to 10% below your base case projections, vacancy runs 200 to 300 basis points higher than modeled, and the exit cap rate is 50 basis points above your base case assumption. The severe downside case pushes those variables further, modeling what happens if the market softens significantly, the renovation takes longer than planned, or interest rates move against you on a floating rate loan.
If the deal still generates acceptable returns in the downside case and doesn't result in a loss of capital in the severe downside case, you have a model that reflects genuine risk discipline. If the deal only works under the most optimistic scenario, that's a signal worth paying attention to before you commit.
Top Picks:

Most multifamily deals that underperform don't fail because of bad luck. They fail because the underwriting was wrong in ways that were predictable and avoidable. Here are the mistakes that show up most consistently in deals that don't deliver what the model promised.
This is one of the most common and most consequential errors in multifamily underwriting. Asking rents, the prices landlords are advertising, are almost always higher than the rents tenants are actually signing leases at. In competitive markets or periods of softening demand, the gap between asking rents and effective leased rents can be significant, sometimes 5% to 10% or more when concessions are factored in.
Always anchor your rent assumptions to actual executed leases on comparable units, not to what properties are advertising on listing platforms. If you can't access actual lease data, apply a concession discount to asking rents to arrive at a more realistic effective rent figure. Building your entire revenue model on asking rents that the market isn't actually clearing is a guaranteed way to overstate projected income from day one.
Sellers present their operating history in the most favorable light possible, which often means expenses that are genuinely too low to sustain once the property changes hands. The three categories where underreported expenses show up most often are property taxes, insurance, and maintenance.
Property taxes in many jurisdictions reset at a higher assessed value after a sale, meaning the seller's current tax bill may be significantly lower than what you'll pay after closing. Get an actual tax reassessment estimate from a local tax consultant before you finalize your underwriting. Insurance premiums have risen sharply across most markets in recent years and should be verified with an actual quote rather than carried forward from the seller's current cost. Maintenance expenses are frequently understated on properties where the owner has deferred repairs or self-manages without properly accounting for their own time.
Exit cap rate is the single assumption that has the most impact on projected returns in most multifamily underwriting models, and it's the one that gets the most wishful thinking applied to it. A sponsor who assumes the property will sell at the same cap rate it was acquired at, or lower, is essentially assuming the market will be as favorable or more favorable at exit than it is today. That's a bet, not an analysis.
Cap rates are influenced by interest rates, investor sentiment, supply and demand dynamics, and broader economic conditions, all of which can shift significantly over a five to seven year hold period. The prudent approach is to underwrite the exit at a cap rate that's 25 to 50 basis points higher than the going-in rate, reflecting the realistic possibility that market conditions at exit won't be as favorable as they are today. If the deal still works at a higher exit cap rate, you have a genuine margin of safety built into the model.
Skipping or minimizing the capital expenditure line in a multifamily underwriting model is one of the fastest ways to produce projections that look great on paper and disappoint in practice. Every apartment property requires ongoing investment to maintain its physical condition and competitive position in the market. Roofs wear out, HVAC systems fail, parking lots need resurfacing, and unit interiors need periodic refreshing to support market rents.
Most institutional underwriters budget $250 to $400 per unit per year in replacement reserves for stabilized properties, with higher amounts for older assets or those with known deferred maintenance. Value-add deals require additional capital expenditure budgets on top of replacement reserves to fund the renovation program. Leaving either of these lines out of the model doesn't make the costs disappear. It just means they show up as surprises during the hold period, eroding cash flow and returns that the model said would be there.
Value-add underwriting models frequently underestimate the income lost during the renovation and lease-up period. Renovating units takes time, and vacant units don't generate income. If your business plan involves renovating 20 units over 18 months, each unit will be offline for some period during that process, and the lost rent during that period needs to be reflected in the model.
The same applies to lease-up assumptions after renovation. Renovated units don't immediately fill at market rents the moment the paint dries. Absorption takes time, and in competitive markets it can take longer than projected. Modeling a realistic lease-up pace rather than assuming immediate full occupancy at target rents is the difference between a projection that survives contact with reality and one that doesn't.
A final mistake worth flagging is the tendency to evaluate deal performance based on gross returns without accounting for the fees, costs, and taxes that reduce what investors actually receive. Acquisition fees, asset management fees, disposition fees, carried interest, and transaction costs all reduce the net return to equity investors relative to the gross return the model generates.
When evaluating a deal as a passive investor, always ask for a model that shows returns net of all fees and carried interest, not just the gross property-level returns. A deal that shows a 2.5x equity multiple at the property level might deliver a 1.9x to 2.1x return to limited partners after fees and promotion are applied. That's still a reasonable result, but it's a meaningfully different number from what the headline projection suggests.
Don’t Miss:
Multifamily underwriting is ultimately an exercise in intellectual honesty. The model will produce whatever numbers you feed it, and the temptation to shade assumptions in the direction that makes a deal look better than it is will always be there, whether you're trying to justify a purchase price you've already fallen in love with or putting together projections to raise capital from investors.
The antidote is discipline. Anchor every assumption to real market data. Stress test the model before you commit, not after. Build in enough conservatism on vacancy, expenses, rent growth, and exit cap rate that the deal still works when things don't go exactly to plan, because they rarely do.