Multifamily underwriting is the work of rebuilding a property's income, expenses and debt with your own assumptions, so you know what the building will really earn under your ownership. You don't accept the seller's net operating income. You test every line against documents and the local market, and you keep only what you can support.
A property manager does this a little differently than a pure investor. We live with the leaks: the unit that sits vacant for six weeks, the insurance renewal, the payroll that never matches the broker's number. Here's how we check a deal.
Start With the Seller's Package
A broker's offering memorandum usually gives you three things: a trailing twelve-month operating statement, a rent roll and a pro forma. Treat them as claims, not facts.
The pro forma is a forecast. The T12 and rent roll are history. Your job is to find the gap between them and decide, line by line, how much of that gap is real. Our guide to the multifamily due diligence checklist covers the documents to request. This article covers what to do with the numbers once you have them.
Lenders do the same thing. Fannie Mae's Multifamily Selling and Servicing Guide (effective February 12, 2025) tells its lenders to use the best available information, including historical performance and anticipated operations, to get to an "Underwritten NCF." It also says to get the prior three years of operating statements where possible, and at minimum the prior full year or the trailing six months annualized. If a lender rebuilds the numbers, so should you.
Multifamily Underwriting Step by Step: How to Underwrite a Deal
How do you underwrite a multifamily deal? Work in order: normalize the T12, verify the rent roll, set market rents, apply vacancy and credit loss, rebuild expenses, add reserves, then test debt and exit. Each step changes the NOI, so do them one at a time and keep notes on why.
Step 1: Normalize the T12
A T12 shows twelve months of actual income and expenses. Before you use it, clean it:
- Tie the income lines to bank deposits where you can. Statements can be accrual-based and flatter than the cash.
- Pull out one-time items: insurance claims, legal settlements, a big repair that won't repeat.
- Look for expenses that were skipped or deferred. A building with almost no repairs spend is not cheap to run; it is behind.
- Check that payroll, utilities and management are all actually on the statement. Owner-operated buildings often leave costs off.
- Compare the last three months to the full year. A trend matters more than the average.
Step 2: Rebuild the Rent Roll
Match the rent roll to the leases. Look at lease start and end dates, rent charged versus rent collected, concessions, and units that are down, model or employee-occupied. Count units, not rows. Then sort by lease expiration: a building where half the leases end in the same two months has a different risk than one with an even spread.
Step 3: Set Market Rents From Comps
Rents in the pro forma should come from comparable properties, not from the broker's hope. Use real comps that match on unit mix, age, finish level and location, and look at what residents actually signed, not just asking rents. A simple comp grid with rents, unit mix and dates keeps the comparison honest.
Then be honest about the gap between in-place rents and market. If you plan to close it through renovations, that is a separate budget and a separate timeline, and the cost belongs in your model.
Vacancy and Credit Loss
Vacancy is the rent you don't get because units are empty. Credit loss is the rent you don't get because residents don't pay. Underwrite them separately, and add concessions as their own line if the market uses them.
Don't copy the seller's vacancy line. Use the building's own history, the comps' occupancy and a turn schedule that makes sense for the unit count. Fannie Mae's guide, for example, subtracts physical vacancy, concessions and bad debt as separate items from gross rental income. That is a good habit for any underwriter: three leaks, three lines.
Other Income and Expenses Line by Line
Other income (parking, laundry, pet fees, utility billbacks, application fees) is where pro formas get optimistic. Use what the building actually collected over twelve months. If you plan to add a fee or a billback program, model it as a later phase, check what local rules and lease terms allow, and don't let it sit in year-one income.
Expenses are where operators earn their keep. Rebuild each one:
- Property taxes. Underwrite the bill you will pay as the new owner, not the seller's. Fannie Mae's guide tells lenders to include an expected increase when a sale would trigger reassessment. Ask the county assessor how it works for the property, and keep a conservative case.
- Insurance. Get a real quote. Don't carry the seller's premium forward.
- Payroll. Staff to the building you'll run. Include taxes, benefits and workers' compensation, and don't forget on-call or after-hours coverage.
- Repairs and maintenance, and turns. Tie them to unit age, turnover and the condition you see on the walk-through.
- Utilities. Use twelve months of bills, and check what residents reimburse.
- Management fee. Model it as a percentage of collected income, using the quote you actually have from your manager. Don't use the seller's owner-managed number.
- Marketing, administration, contracts. Small lines, but they add up across a year.
Every one of these lines flows straight into NOI, so a small miss here becomes a large miss in value.
Reserves and the Capital Plan
Reserves are a real cost even when nobody forces you to fund them. Fannie Mae's guide requires reserve expense in underwritten net cash flow whether or not the escrow is funded, with a minimum of $200 per unit a year or more where required by its reserve rules. That is the lender's floor, not your answer. Your number should come from the condition of the roofs, paving, mechanical systems and appliances.
Keep a separate capital budget for the work you plan to do at purchase: deferred maintenance, safety items and any value-add scope. Fold it into your sources and uses so the deal isn't underfunded on day one.
Debt Terms and Debt Service Coverage Ratio
Once you have your own NOI, test it against the loan. The debt service coverage ratio divides net operating income by annual debt service, and lenders size loans so that it clears a minimum.
Check a few things:
- Is the rate fixed or floating, and what happens if it resets?
- How long is the interest-only period, and what does payment look like after it ends?
- Does the lender size on your NOI, theirs or the seller's?
- Are there reserves, holdbacks or prepayment terms that change your cash needs?
Run the DSCR on your normalized NOI, not the pro forma. If the deal only works on the seller's numbers, it doesn't work.
Exit Cap Rate and Sensitivity
The exit is the last assumption and often the loosest. Investors usually reason about exit cap rate by asking what a buyer would pay for the building you expect to own, not what the market pays today. A common habit is to use an exit cap rate at or above the going-in rate, but that is a habit, not a rule, so test a range.
Sensitivity testing is the part many models skip. Change one assumption at a time and watch what happens:
- Rents grow slower than your plan.
- Vacancy runs higher for a year.
- Taxes reset higher after the sale.
- Insurance jumps at renewal.
- The exit cap rate comes in wider.
If one bad assumption breaks the deal, you've learned where your risk is.
The Assumptions Table: What to Check
Use this as a working list. It gives no benchmark numbers on purpose: the right figure depends on the building, the submarket and the date.
| Assumption | What to check |
|---|---|
| Gross potential rent | Unit count and mix tied to the rent roll and leases |
| Market rent | Signed rents on true comps, not asking rents |
| Loss to lease | Gap between in-place and market rents, and how fast it closes |
| Vacancy | Building history, comp occupancy and a realistic turn schedule |
| Concessions | What the market is offering and what the building offered last year |
| Credit loss | Aged receivables and collections, not the seller's summary |
| Other income | Twelve months of actual collections, and what is lease-permitted |
| Property taxes | The post-sale bill and the county's reassessment rules |
| Insurance | A current quote for this building and its claims history |
| Payroll | Staffing plan, taxes, benefits, workers' comp, on-call |
| Repairs and turns | Unit condition, turnover rate and vendor pricing |
| Utilities | Twelve months of bills and what residents reimburse |
| Management fee | A written quote, as a percent of collected income |
| Reserves | Property condition, not only the lender minimum |
| Capital budget | Scope and bids for day-one and value-add work |
| Debt terms | Rate type, interest-only period, sizing basis, reserves |
| Exit cap rate | A range, with the reasoning for each case |
If you want a structure to drop these into, our multifamily underwriting model lays them out in a spreadsheet.
An Illustrative Example of the Gap
Here is a made-up example. These numbers are illustrative only, not a benchmark and not from a real deal.
Say a broker's pro forma shows a 40-unit building with rents rising to $1,600 a unit and very low vacancy. The T12 shows rents averaging $1,450, two units down for repairs, and an insurance bill that is about to renew. You run the checks:
- The pro forma rent is $150 above in-place. Comps support part of that gap, and closing it needs renovations that aren't in the budget.
- The two down units are not vacancy, they are a repair problem. You add the cost to fix them.
- You price a new insurance quote and a higher post-sale tax bill.
None of those changes is dramatic on its own. Together they can move your NOI and your DSCR enough to change what you should offer. That is the whole point of underwriting.
Where a Property Manager Fits
Many mistakes in broker pro formas are operating mistakes: rents the market won't pay, turn costs too light, payroll too thin, a management fee that doesn't match reality. A manager who works in the market can catch those in an afternoon.
If you want your numbers reviewed by the people who will run the building, our team does multifamily financial modeling for owners and investors in Reno, Sparks, Carson City and Lake Tahoe. We don't invent returns. We check the assumptions, show you the range and tell you where the risk sits.
This article explains concepts and isn't tax, legal or investment advice. Talk with your attorney, tax advisor and lender about your own deal.
