Overview
Understanding how to underwrite a multifamily deal is a core competency for any CRE investment professional. Done well, underwriting tells you whether a deal pencils at the ask, where the risk is concentrated, and what has to be true for the investment to hit its return hurdles. Done poorly, it produces a model that looks clean but breaks down under the first real scenario.
The process follows a logical sequence: get the documents, verify the numbers, build the model, and pressure-test the thesis. This guide covers each step in the order that professional analysts actually work through them, with the specific details that matter at each stage.
A note on time: manual multifamily underwriting typically takes 8 to 12 hours for a first-pass model on a 100 to 200 unit property. That timeline is compressing as AI-assisted platforms automate the data extraction and model-building steps, but the analytical judgment required at each step remains the same.
Step 1: Review the Offering Memorandum
The OM is the seller's narrative. Your job in this step is to separate the verifiable facts from the broker's forward projections, and to identify what you need to confirm independently before the model can be trusted.
Work through the OM with these questions in mind:
- What is the stated NOI, and how was it calculated? Brokers often present a "pro forma NOI" based on market rents and optimistic occupancy. Note every line item that deviates from trailing actuals.
- What is the unit mix and total unit count? Confirm these match the rent roll you will receive separately. Discrepancies at this stage are a flag.
- What capital has been spent, and what remains? Review the renovation summary closely. Sellers routinely overstate rehab quality and understate remaining deferred maintenance.
- What are the lease terms and expiration schedule? A property with 60% of leases rolling in year one carries meaningfully different risk than a stabilized book.
- What debt is in place? Note existing financing, assumability, prepayment penalties, and any associated reserve accounts. For value-add deals, confirm whether agency financing is available at the projected stabilized basis.
Document every assumption the broker has made in the pro forma. You will revisit each one when you set your own assumptions in Step 4.
Step 2: Analyze the Trailing Twelve-Month Financials
The T12 (trailing twelve months) is the property's operating statement for the most recent twelve-month period. It is the single most important document in multifamily underwriting because it reflects what the property actually earned and spent, not what the seller projects it will earn.
T12 reconciliation involves several critical checks:
- Normalize one-time items. Identify any non-recurring income or expense. A one-time insurance recovery, a legal settlement, or deferred maintenance catch-up will distort NOI if left in. Add them back or remove them before establishing your baseline.
- Verify revenue line by line. Gross potential rent, physical vacancy loss, concessions, bad debt, and other income should each be traceable. Total effective gross income (EGI) is what you are underwriting from, not GPR.
- Scrutinize controllable expenses. Repairs and maintenance, payroll, and management fees are the most commonly manipulated line items in seller-provided T12s. Compare each against market benchmarks for your property class and submarket.
- Calculate trailing NOI with your own numbers. Do not accept the seller's stated NOI. Run your own EGI minus your own operating expenses, and note the gap versus what the broker is presenting.
- Cross-reference with the rent roll. The rent roll's in-place rent totals should roughly reconcile with the T12's gross potential rent. Significant variances require explanation.
For complex properties or those with messy books, T12 reconciliation can take two to three hours manually. Platforms like AcquiOS automate the extraction and normalization step, flagging anomalies for analyst review rather than requiring line-by-line manual entry.
Step 3: Process the Rent Roll
The rent roll is a unit-by-unit schedule of current tenants, lease terms, in-place rents, and unit types. Rent roll analysis is where you quantify the gap between current revenue and market potential, and where a significant portion of underwriting risk lives.
A properly processed rent roll gives you the following data for each unit type in the property:
| Unit Mix | In-Place Rent | Market Rent | Loss to Lease | Occupancy |
|---|---|---|---|---|
| Studio (40 units) | $1,050 | $1,150 | $100/unit | 95% |
| 1BR/1BA (80 units) | $1,320 | $1,475 | $155/unit | 93% |
| 2BR/1BA (50 units) | $1,680 | $1,850 | $170/unit | 91% |
| 2BR/2BA (30 units) | $1,920 | $2,100 | $180/unit | 90% |
Loss to lease is the dollar gap between in-place rents and current market rents. It represents rent upside as leases turn over, but requires careful underwriting: you need to verify that the cited market rents are real (pull your own comps), and you need to account for the time and costs required to achieve them.
Key rent roll checks:
- Flag any units with above-market in-place rents. These units carry rolldown risk on renewal.
- Identify the lease expiration schedule. Concentrated near-term expirations create both opportunity and cash flow risk.
- Note any MTM (month-to-month) leases. These are immediately convertible but also immediately at risk of vacancy.
- Check for any concessions listed on individual leases. They reduce effective rent below the stated in-place figure.
- Compare total scheduled rent on the roll against the T12 GPR line. Reconcile any gap.
Step 4: Set Market-Calibrated Assumptions
This is the step where underwriting judgment matters most. Your assumptions will drive the model's output more than any other input, and they need to be defensible to both an investment committee and a debt lender.
Core assumption categories for multifamily:
- Rent growth: Set year-by-year rent growth based on submarket fundamentals. Source your own comps from CoStar, CBRE, or broker market reports. Do not accept the broker's rent growth assumptions without independent verification. A 3% flat assumption is often more defensible than a custom hockey stick.
- Economic vacancy: Model physical vacancy (units unoccupied) and loss to lease separately. For a stabilized asset, 5 to 7% economic vacancy is typical in most markets. Value-add properties in lease-up may require 10 to 15% in the first 12 to 18 months.
- Expense ratio: For class B/C multifamily, total operating expenses typically run 40 to 50% of EGI. For class A, 35 to 45%. Management fees should be 4 to 6% of EGI depending on property size. Model reserves separately at $250 to $400 per unit per year for a B-class asset.
- CapEx: Distinguish between renovation capital (budgeted, time-limited) and ongoing replacement reserves. For value-add deals, model each renovation tranche with its associated timeline and rent premium capture rate. Be conservative on lease-up velocity.
- Exit cap rate: Apply a 25 to 50 basis point spread to today's going-in cap for your exit assumption. Investors who underwrite exits at compressed cap rates are pricing in risk they are not acknowledging. For a 5-year hold, your exit cap assumption is the single most consequential number in the model.
- Financing: Current agency debt (Fannie/Freddie) terms for stabilized multifamily are typically DSCR 1.20 to 1.25x minimum. For floating-rate bridge, model the full rate range including a stressed rate scenario. I/O periods and extension options should be explicitly modeled, not assumed away.
Step 5: Build the Discounted Cash Flow Model
The DCF is the core deliverable of multifamily underwriting. It projects property cash flows over the hold period, discounts them to present value, and produces the return metrics your IC will use to make a decision.
Structure the DCF in this order:
- Gross Potential Rent (GPR): Unit count times market rent by type, grown annually by your rent growth assumptions.
- Vacancy and credit loss: Subtract physical vacancy, loss to lease, concessions, and bad debt to arrive at Effective Gross Income (EGI).
- Other income: Add parking, laundry, pet fees, RUBS, and any other ancillary income streams. Model these conservatively; they are frequently inflated in OM pro formas.
- Operating expenses: Apply your normalized T12 expense baseline, grown at CPI (typically 2.5 to 3.5%). Model management fees as a percentage of EGI, not a fixed dollar amount.
- Net Operating Income (NOI): EGI minus total operating expenses. This is your key operating metric at every point in the hold period.
- Debt service: Apply your financing terms to calculate annual debt service. Calculate DSCR for each year. Flag any year where DSCR falls below your lender's covenant.
- CapEx: Model renovation spend by tranche. For value-add deals, this typically front-loads negative cash flow in years one through two.
- Net Cash Flow: NOI minus debt service minus CapEx. This is what accrues to equity each period.
- Exit proceeds: Apply your exit cap rate to year-n+1 NOI, subtract disposition costs (typically 1.5 to 2.5% of gross sales price), and repay the outstanding debt balance. Net proceeds go to equity.
AcquiOS's DCF generation feature builds this structure automatically from the extracted OM, T12, and rent roll data, pre-populating your existing Excel template rather than forcing a new format on your team.
Step 6: Calculate Return Metrics
With the DCF built, your model produces the three primary return metrics that multifamily investors use to evaluate deals:
- IRR (Internal Rate of Return): The annualized return on equity invested, accounting for the timing of all cash flows including the exit. For core-plus multifamily, target IRRs typically range from 8 to 12%. Value-add deals typically target 13 to 18%. The IRR is highly sensitive to hold period and exit cap rate. Run the IRR calculation at multiple exit scenarios before presenting to IC.
- Cash-on-Cash Return: Annual net cash flow divided by total equity invested. This measures current yield. For stabilized assets, 5 to 7% cash-on-cash is typical in today's market. Value-add deals often produce negative or minimal cash-on-cash in years one through two while the renovation program runs, with a ramp to 6 to 9% post-stabilization.
- Equity Multiple: Total equity returned (including exit proceeds) divided by equity invested. A 5-year hold targeting a 1.8x to 2.2x equity multiple on a value-add deal is a common benchmark. Do not conflate equity multiple with IRR. A 2.0x multiple over 10 years is materially inferior to 2.0x over 5 years.
Present all three metrics in your IC model. Some committees weight IRR; others prioritize equity multiple as a simpler measure of capital preservation and growth. Know your audience.
Step 7: Stress Test the Model
A model that only works at base case is not a model you can defend to an investment committee. The stress test is where you prove you understand the risks and have sized them appropriately.
Run at minimum three scenarios in addition to base case:
- Downside scenario: Rent growth flat for two years, physical vacancy 200 to 300 bps above base, exit cap 50 bps higher, renovation budget 15% over. What is the IRR, and does the deal still service its debt?
- Upside scenario: Rent growth at the top of the market range, faster lease-up, exit cap in line with today. This quantifies the ceiling on the investment and is useful for IC dialogue on risk-adjusted returns.
- Rate stress (for floating-rate debt): If you are using bridge financing, model the debt service at 200 bps above your base rate assumption. Does the property still cash-flow positive? Does the DSCR breach a loan covenant?
Sensitivity tables are more useful than individual scenarios. Build a two-axis sensitivity showing IRR across combinations of exit cap rate (rows) and rent growth rate (columns). This immediately shows your IC where the deal breaks and what has to go wrong for returns to fall below the hurdle.
The most common mistake at this stage is stress-testing the wrong variables. Exit cap rate and rent growth matter more than most teams model. Lease-up timing on value-add deals is consistently underestimated. And the cost of capital on bridge debt in a rising rate environment has been the single biggest driver of deal distress in recent years. Weight your scenarios accordingly.
Frequently Asked Questions
How long does it take to underwrite a multifamily deal?
Manual underwriting on a 100 to 200 unit multifamily property typically takes 8 to 12 hours for a first-pass model: 2 to 3 hours for T12 reconciliation and rent roll processing, 2 to 3 hours for the DCF build, and 1 to 2 hours for assumption research and stress testing. Revisions for IC add additional time. AI-assisted platforms like AcquiOS compress the data extraction and model-building steps to under 35 minutes, allowing analysts to focus on the judgment-intensive steps.
What return metrics matter most in multifamily underwriting?
The three core metrics are IRR, cash-on-cash return, and equity multiple. IRR is the most widely cited and accounts for the timing of all cash flows. Cash-on-cash measures current yield and matters more for income-oriented investors. Equity multiple is the simplest measure of how much capital you are getting back and is often the clearest metric for IC discussions. Most investment committees want to see all three, and they tell different parts of the story.
What is the difference between economic and physical vacancy?
Physical vacancy is the percentage of units that are unoccupied at a given time. Economic vacancy is a broader measure that includes physical vacancy plus concessions, loss to lease, bad debt, and any other reductions between gross potential rent and actual collected revenue. Economic vacancy is the right metric for underwriting because it captures all the ways revenue falls short of theoretical potential, not just empty units. A property with 95% physical occupancy can still have 10 to 12% economic vacancy if concessions and loss to lease are high.
Can I use AI to underwrite multifamily deals?
Yes, and the technology has matured significantly. AI platforms purpose-built for CRE underwriting can extract data from offering memorandums, T12s, and rent rolls, populate your existing Excel model, and flag anomalies for analyst review. The key is that AI handles the time-consuming data extraction and model population steps while the analyst retains control over assumptions and final judgment calls. AcquiOS's rent roll analysis and T12 reconciliation features are specifically built for multifamily workflows, outputting directly into the team's existing template.