What Underwriting Actually Means

Underwriting a multifamily deal is the process of answering one question: at this price, with these assumptions, does this property generate enough income to justify the capital and the risk?

That sounds simple. It is not. Most retail investors confuse "running the numbers" with underwriting. Running the numbers means plugging inputs into a calculator and accepting the output. Underwriting means stress-testing every assumption, identifying which ones break the deal if they go wrong, and deciding whether you are being compensated for the risks you are taking.

This article walks through the institutional approach. Not the BiggerPockets version, not the broker pro forma version. The version that institutional analysts use to recommend or reject deals to investment committees.

Step 1: Verify the Income

Every deal starts with the rent roll. Before you trust a single number, verify the following:

In-place rents vs. market rents. The broker will show you potential rents. Your job is to figure out actual rents being paid right now. Pull the T-12 (trailing 12 months) and divide actual collected rent by occupied units. That number is your real starting point, not the proforma.

Loss to lease. Compare in-place rents to market rents for the same unit type in the same submarket. If in-place rents are 15%+ below market, that's loss to lease. It looks like upside on paper, but capturing it takes 18-24 months as leases roll. Underwrite to actual rents for the first year, not proforma rents.

Concessions and bad debt. Ask for the rent collection report. Are tenants getting one month free? Are there delinquencies the seller isn't disclosing? Subtract concessions and bad debt from gross potential rent before you calculate effective gross income.

Vacancy. Use 5% as a baseline for stabilized multifamily. If the property is below 90% occupied, find out why before underwriting any growth. Use the actual current occupancy, not market average.

Step 2: Audit the Expenses

The seller's pro forma will always understate expenses. Always. Your job is to model expenses as they will actually run under your ownership.

Property tax reassessment. Most counties reassess property taxes upon sale. The current owner might be paying $30K based on a 2018 valuation. After you buy, the assessor revalues based on your purchase price, and your tax bill could double. Model this in year 1, not year 5.

Insurance. Insurance has tripled in some markets since 2022 due to climate exposure and reinsurance costs. Get a real quote, not a "use the seller's number" estimate. Florida, Texas, and California are especially exposed.

Property management. If you are not self-managing, model 4-7% of effective gross income for management fees, depending on unit count. Don't assume you'll do it yourself "to save money." Either pay for management or build the labor cost into your numbers.

Capital reserves. Multifamily institutional standard is $250-350 per unit per year for capex reserves. This is for replacing roofs, HVACs, parking lots, exteriors. Below the line in an Excel model, but very real in actual cash flow. NCREIF data shows CapEx averaging roughly 30% of NOI annually across property types. A 7% cap rate deal with typical CapEx generates a true distributable yield closer to 5%. If you're quoting cap rate to your investors without mentioning CapEx, you're overstating returns by 100-200 basis points.

Repairs and maintenance. Separate from capex. R&M is the day-to-day stuff: turnover costs, plumbing calls, painting between tenants. Budget $400-600 per unit per year minimum.

If your expense ratio is below 35% of effective gross income, you have likely understated something. Most stabilized multifamily runs 38-50% depending on age, amenities, and location.

One more expense trap: lease type. Multifamily leases are the shortest in CRE (month-to-month or 12-month terms). This means tenants reprice to market continuously, which is great when rents are rising and painful when they're falling. Unlike NNN industrial leases where tenants absorb operating costs, multifamily landlords bear all operating expense risk directly. Every insurance increase, every tax reassessment, every utility rate hike flows through to your NOI in real time.

Step 3: Stress Test the Debt

Most deals look great until the debt service hits. Your DSCR (debt service coverage ratio) is the single most important number for whether you can finance the deal at all.

DSCR thresholds: Lenders typically require minimum 1.25x DSCR. Below 1.0x, the property can't cover its own debt service. Below 1.20x, you'll struggle to find financing at conventional rates. Above 1.30x is comfortable.

Debt yield: This is the metric institutional lenders increasingly care about. NOI divided by loan amount. A typical lender floor is 9-10%. Below 8% is a hard pass for most institutional lenders. Why? Debt yield doesn't depend on interest rates. It tells the lender how quickly they could recover their loan if they had to take the property back.

Stress test rates. Don't underwrite to today's rate as if it will hold forever. Run a scenario where rates are 150 basis points higher at refinance. If your DSCR drops below 1.0x in that scenario, you have refinancing risk. The Fed's DFAST methodology uses scenarios up to a 40% valuation decline.

Here's the sobering context: academic research on approximately 18,000 commercial mortgages (1972-1997) found that 1 in 6 defaults over its lifetime, with defaults peaking around year 6. Loans originated at peak valuations (1986 vintage) defaulted at a 28% rate. The lender's DSCR floor isn't conservative; it's calibrated against a world where this happens regularly.

Interest-only periods. If the loan offers an IO period, recognize that your cash flow during IO is artificial. Once amortization starts, cash flow drops. Model both periods separately.

Step 4: Check the Leverage Direction

This is the metric most retail investors miss: positive vs. negative leverage.

If your unlevered cap rate is 6% and your cost of debt is 7%, every dollar you borrow makes your return worse. That's negative leverage. You are paying the lender more than the property earns.

The check: cap rate minus loan constant (annual debt service divided by loan amount). If positive, leverage amplifies your returns. If negative, leverage destroys them. Use the loan constant, not the stated interest rate. For the full framework, read Positive vs Negative Leverage. For amortizing loans, the loan constant exceeds the interest rate because it includes principal repayment. A 6.75% rate on a 30-year amortizing loan has a loan constant near 7.8%.

Negative leverage isn't always disqualifying. Sometimes a value-add deal carries negative leverage during the renovation period because the in-place NOI hasn't caught up yet. But you need a clear thesis for how you get to positive leverage by year 2 or 3. Without that thesis, you are paying yourself to lose money.

Step 5: Project the Hold and the Exit

The going-in numbers tell you whether the deal works on day one. The hold period numbers tell you whether the deal works as an investment.

Rent growth assumptions. Don't use 5% rent growth because Yardi forecasts it. Use the trailing 5-year actual rent growth in your specific submarket. If you are projecting growth above the historical actual, you need to explain why.

Expense growth. Always model expense growth at 1-2% above rent growth. Insurance and property tax growth historically outpace rent growth. If you model both at 3%, you are being optimistic.

Exit cap rate. This is where deals get killed. Most pro formas assume the property sells at the same cap rate it was purchased at. That's a flat assumption, and flat assumptions are wrong. The institutional approach: model exit cap at 50-100 basis points above going-in cap. If the deal still works at that exit, you have a margin of safety. If it only works at a flat or compressed exit cap, you are betting on cap rate compression rather than operating performance.

Sensitivity analysis. Run your IRR across a 5x5 matrix of exit cap rates and rent growth scenarios. How many cells produce an acceptable return? If only 5 of 25 do, the deal requires perfect execution. If 18 of 25 do, you have real margin.

The Bias Check for Multifamily Underwriting

Before you commit to a deal, ask yourself:

  1. Have I verified actual in-place rents and expenses, or am I using broker pro forma numbers?
  2. Have I modeled property tax reassessment, current insurance quotes, and proper capital reserves?
  3. Is my DSCR comfortable today AND at a 150 basis point higher refinance rate?
  4. Am I in positive leverage from day one, or am I betting on getting to positive leverage later?
  5. Does my exit cap rate add risk premium, or am I assuming flat or compressed cap rates?
  6. How many cells of my sensitivity matrix produce my target IRR?

If you can answer all six honestly and the deal still works, you have a real opportunity. If not, you have a hope.

The Mental Model: Three Filters, In Order

Institutional buyers don't evaluate every aspect of a deal simultaneously. They run three filters in sequence, and they stop as soon as a filter fails.

Filter 1: Does the income hold up? Verify in-place rents against the T-12. Compare to market comps. If the pro forma rents are materially above actuals with no documented plan to close the gap, stop. The income assumption is the foundation. If it is wrong, every metric built on top of it is wrong.

Filter 2: Does the debt work? Calculate all three DSCRs (going-in, stabilized, stressed). Check the leverage direction (cap rate vs. loan constant). If the deal is negatively levered with no credible path to positive, or if the stressed DSCR falls below 1.0x, stop. The financing will eventually force your hand.

Filter 3: Does the exit make sense? Model exit cap rate 50 to 100 basis points above going-in. Run a 5x5 sensitivity matrix of exit cap vs. rent growth. Count the cells that produce your target IRR. If fewer than half the cells work, you need perfect execution to hit your target. That is not investing.

Most deals fail at Filter 1. The ones that pass all three are the ones worth spending time on diligence, legal, and inspections. Everything before Filter 3 should take less than an hour.

Where to Go Next

  • Audit the pro forma line by line. Read How to Read a Pro Forma for the PGI-to-PBTCF cascade with a worked broker vs. buyer comparison.
  • Understand the cap rate you're paying. Read What Is a Good Cap Rate to decompose the growth bet embedded in the price.
  • Stress test the financing. Read DSCR Explained for the three-DSCR framework and the IO cliff.
  • Avoid the common traps. Read 8 Underwriting Mistakes for the full checklist of errors that inflate returns.

Sources

  • Geltner et al., "Commercial Real Estate Analysis and Investments," 2nd Edition (2007), Chapters 11, 13, 18: Cash Flow Proformas, Leverage, Mortgage Underwriting
  • Esaki et al., "Commercial Mortgage Defaults: An Update" (2002)
  • Krewson-Kelly and Thomas, "The Intelligent REIT Investor" (2016), Chapter 4: Lease Structures
  • Adventures in CRE, "Capital Reserves Case Study"
  • CBRE U.S. Cap Rate Survey, H2 2025
  • NBER Working Paper 31970, "Monetary Tightening, Commercial Real Estate Distress, and US Bank Fragility" (2023)
  • Federal Reserve DFAST stress testing methodology
  • NCREIF Property Index: CapEx empirical data (1978-2004)