Deal Underwriting for Investors: From Rent Rolls & T12s to ARV & MAO

Underwriting is the discipline of turning a pile of imperfect inputs - a listing photo, a contractor's ballpark, a broker's cash-flow statement scanned into a PDF - into a defensible number you are willing to sign under. Do it well and you know your margin before you are committed. Do it poorly and you learn your margin after closing, which is the expensive way to learn it.

This guide walks both sides of the market. First the single-family and wholesale mechanics - comps, ARV, the 70% rule, rehab, and the assignment spread. Then the multifamily stack - the T12, the rent roll, net operating income, cap rate, and the debt-service coverage your lender will insist on. Then the part no one puts on a webinar slide: the operational grind of getting messy inputs into a model fast enough to actually win the deal.

What real estate deal underwriting is - and isn't

Underwriting is not an appraisal. An appraiser certifies value for a lender using a rulebook; you are estimating value and risk from the buyer's chair, for your own account, with your own business plan baked in. The same building is worth different numbers to a flipper, a buy-and-hold landlord, and a syndicator, because each one does something different with it.

So learning how to underwrite a real estate deal is really learning to answer two questions quickly and conservatively: what will this asset be worth once you have executed your plan, and what is the maximum you can pay today to hit your target return with a margin for the things that will go wrong. Everything below is machinery for producing those two numbers.

The single-family side: comps, ARV, and the 70% rule

On a house, value starts with comparable sales. Pull three to six properties that recently sold (not listed) nearby, similar in size, age, and style, and adjust for the obvious differences - a garage, an extra bath, a finished basement. The number you land on is the after-repair value (ARV): what the house will fetch once it is fixed and sold in ordinary condition. ARV is the anchor for everything else; get it wrong and no amount of clever math downstream saves you.

From ARV, most investors work backward with the 70% rule to a maximum allowable offer (MAO). The rule says you should pay no more than 70% of ARV, minus repairs. Written as a formula, the version lenders and educators publish is straightforward:

MAO = (ARV × 0.70) − estimated repairs − fixed costs The 70% rule, as framed in Chase's homebuyer education: offer no more than 70% of ARV minus costs (Chase).

The missing 30% is not profit - it is a cushion that absorbs holding costs, closing costs, selling commissions, and the estimating errors that are guaranteed on a house you have not yet opened up. Take a $300,000 ARV with $50,000 of repairs and $10,000 of fixed costs: MAO = (300,000 × 0.70) − 50,000 − 10,000 = $150,000. Offer more than that and you are spending your margin before you own the asset. The 70% figure is a guideline, not a law; in supply-starved markets flippers stretch to 75% or higher, and on cheaper houses the fixed-dollar costs eat a bigger share, so the percentage tightens.

Estimating rehab without kidding yourself

Repairs are where new investors bleed, because optimism is free and drywall is not. The reliable habit is to estimate in buckets rather than one hopeful lump: a cosmetic tier (paint, flooring, fixtures, landscaping), a systems tier (roof, HVAC, electrical, plumbing, foundation), and the expensive surprises that only appear once walls are open. Price each bucket from real local numbers, then add a contingency of 10–20% on top - not because you are careless, but because the house is old and honest.

Two disciplines separate operators who survive from those who don't. First, walk the property or send someone who can, because a photo hides the water stain. Second, tie the scope to the exit: a rental refresh and a flip finish are different budgets for the same house, and pretending otherwise is how a deal that pencils on paper loses money in the field.

The wholesale spread: where your margin lives

Wholesalers rarely take title. You put a property under contract at a price, then assign that contract to an end-buyer - usually a flipper or landlord - for a fee. Your margin, the spread, is what that end-buyer will pay minus your contract price minus your assignment fee. The trap is that the end-buyer is running the exact 70%-rule math above, so their ceiling is their MAO. If you contract a house at a price that sits above the end-buyer's MAO once your fee is added, there is no spread and no deal - only a contract you cannot move. Good wholesaling is therefore just disciplined underwriting done one seat earlier in the chain: you have to know the investor's numbers to protect your own. Much of that starts upstream, in how investors generate motivated-seller leads priced below what a retail buyer would pay.

The multifamily side: reading a T12 and a rent roll

Cross into small and mid-market multifamily and the vocabulary changes. Value no longer comes from what the neighbor's house sold for; it comes from the income the building produces. Two documents drive everything, and a broker will hand you both in the offering package.

The T12 - trailing twelve months - is the property's actual income and expense statement for the last year, month by month. It is the closest thing to ground truth about how the building really performs, as opposed to how a marketing pro forma wishes it performed. The rent roll is a unit-by-unit snapshot: each unit, the tenant, lease start and end, current rent, market rent, and status. Read together, the T12 and rent roll tell you what the building earns today and where the slack is: units renting below market (loss-to-lease), heavy concessions, delinquency, or a stack of month-to-month leases that signal turnover. Below-market rents are not a red flag - for a value-add buyer they are the whole thesis.

From gross income to NOI: normalizing the numbers

The engine of multifamily underwriting is net operating income (NOI), and it is built in a fixed order. Start with gross potential rent, subtract vacancy and credit loss to get effective gross income, then subtract operating expenses - property taxes, insurance, management, repairs and maintenance, utilities, payroll, and reserves. What remains is NOI. Crucially, NOI excludes your mortgage, capital improvements, and depreciation; it measures the property's performance, not your financing.

The word that matters most here is normalize. Never underwrite the seller's numbers as given. Re-underwrite property taxes to what they will be after the sale triggers a reassessment, not the seller's grandfathered basis. Put insurance at a real quote in today's market. Charge management at a market rate - typically a percentage of collected income - even if the seller "self-manages" for free, because you will pay someone eventually. Add capital reserves per unit. A broker's pro forma tends to inflate income and shave expenses; normalization is how you drag it back to reality before it costs you.

Cap rate, value, and the DSCR your lender cares about

Once you trust the NOI, two ratios do the heavy lifting. The capitalization rate is annual NOI divided by price; rearranged, value equals NOI divided by the market cap rate. Cap rates are set by the market and vary sharply by geography and asset quality. Across the 30 most active U.S. multifamily markets, average cap rates ran about 5.04% in 2025, but the spread was wide - from roughly 3.88% in San Francisco to 6.27% in Fort Lauderdale, per Yardi Matrix data (CRE Daily). A single national number is useless for pricing a specific building; you underwrite to the cap rate of its submarket and class.

The second ratio belongs to your lender. Debt-service coverage ratio (DSCR) is NOI divided by annual debt service (principal plus interest). It tells the bank whether the building's income covers its loan payments with room to spare. Most commercial lenders start at a minimum around 1.25x, and multifamily conventionally sits right at that 1.25x line (Commercial Real Estate Loans). Practically, DSCR sizes your loan: the lender will lend only up to the amount whose annual payment your NOI can cover at their required ratio. On $70,000 of NOI at a 1.25x minimum, the maximum annual debt service the bank will underwrite is $70,000 ÷ 1.25 = $56,000 - and that constraint, not the sticker price, often decides how much you can actually borrow.

The value-add math

Because value equals NOI divided by cap rate, every recurring dollar you add to NOI multiplies. This is the entire logic of value-add. Raise rents $100 a month across 40 units and you have added $48,000 of annual NOI; at a 5.5% cap rate, that is roughly $873,000 of created value ($48,000 ÷ 0.055) - from a single line item. The same lever works on the expense side: trim a padded management contract or a mispriced insurance policy and the savings capitalize just like new rent. The discipline is to underwrite the value-add as a plan you can actually execute on a timeline, not to pay the seller today for the NOI you hope to create over the next three years.

The real bottleneck: turning a broker's PDF into a model

Here is the part that decides more deals than any formula above: the math is not the hard part - the data wrangling is. A broker sends a 40-page offering memorandum, a T12 as a scanned image, and a rent roll locked inside a formatted spreadsheet. Before you can even begin to underwrite, someone has to re-key all of it into a model, reconcile the rent roll against the T12, and normalize a dozen expense lines by hand. On a single-family deal it is a contractor's scrawled estimate and a pile of comps. Either way, hours disappear into transcription before a single decision gets made.

And speed is not a nicety in this business - it is the edge. The best deals draw multiple offers within days, and the investor who can produce a credible number the same afternoon is the one who gets the letter of intent taken seriously. The investor still hand-typing a rent roll a week later is bidding on a property that is already under contract. That same pressure lands on the phone, where a rep often has to move a seller from comps to a defensible maximum allowable offer live, in the middle of the call. That is what real-time call coaching is built for - it listens to the live seller call and prompts the rep through the comps-to-MAO conversation as it happens, which is why our parent company backs work in this space, alongside broader shifts in how AI is changing real estate investing. Whatever you adopt, the principle holds: the faster you can turn documents into a defensible number, the more deals you get to say yes to before someone else does.

A worked example, end to end

Take a ten-unit building, all occupied, in-place rents of $1,150 against a market of $1,300. The listing broker's pro forma shows the market-rent future and a lean expense load; your underwriting prices the building on what it earns today, with normalized costs. The two columns below are the same asset seen from two chairs.

The $800,000 gap between the two value lines is not a rounding error - it is the sum of an inflated rent assumption and three missing expense lines. You would offer near the $1.29M your in-place NOI supports, then earn the upside by executing a rent and expense plan over the next couple of years to move NOI toward the pro forma. Financing follows the same conservative NOI: at a 1.25x DSCR, your $71,112 supports about $56,890 of annual debt service, which - not the seller's asking price - sets your real loan ceiling. The seller's number is a story about the future; your number is a price you can defend today.

A checklist before you sign

Strip both sides of the market to a sequence you can run on any deal:

1. Establish value - comps and ARV on a house; normalized NOI and a submarket cap rate on a building.
2. Underwrite the costs - bucketed rehab plus contingency, or a re-underwritten expense stack with reserves.
3. Back into a price - the 70% rule and MAO on single-family; NOI ÷ cap rate on multifamily.
4. Check the financing - does NOI clear the lender's DSCR, and does the loan you can get make the deal work?
5. Price the plan, not the dream - pay for in-place performance; earn the value-add yourself.
6. Move fast - a credible number produced today beats a perfect one produced next week.

None of this requires a proprietary spreadsheet or an MBA. It requires being conservative where it counts, honest about repairs and expenses, and quick enough to act - because in a market where the same building is worth different numbers to different buyers, the disciplined underwriter is simply the one least likely to overpay. For the wider context around today's opportunities, see our field guide to real estate investing in 2026 and the state of the U.S. distressed-property market.