How to Underwrite a Multifamily Deal: A 7-Step Framework

Most multifamily deals die at the first screen. That is the point of a first screen, but it also means the expensive part of underwriting is doing it slowly on deals that were never going to work.

What follows is the framework we built MultiScreen around: seven steps, in order, each one designed to kill a bad deal as early as possible and to leave you with a defensible number on the ones that survive.

Step 1: Rebuild NOI from the T12, not the proforma

The offering memorandum's proforma reflects the seller's business plan under the seller's assumptions. It is advocacy, not analysis, and starting from it means inheriting every assumption you have not yet examined.

Start instead from the trailing twelve and rebuild net operating income line by line. Three adjustments are almost always required and almost always missing:

• Property taxes at your basis. Most jurisdictions reassess on sale. A seller who has owned since 2015 is carrying a tax load you will never see.
• Management fee at market. Typically 3% to 4% of effective gross income, applied even when the seller self-manages and reports no fee.
• Replacement reserves. $250 to $300 per unit per year is the common agency underwriting standard.

Then look at the monthly detail rather than the annual column. Compare the trailing three and trailing six months, annualized, against the full twelve. A widening gap means the run rate has moved and the annual figure is already stale.

Step 2: Verify the revenue against the rent roll

The T12 tells you what was collected. The rent roll tells you why, and whether it repeats.

Reconcile the two: annualize the rent roll and check that it ties roughly to trailing collections. When it does not, the gap is the question worth asking. Then look for the things that occupy a unit without paying for it, which are invisible in an occupancy percentage:

• Down units awaiting renovation
• Employee and model units
• Units on concession, where the face rent overstates the effective rent

This is the difference between physical and economic vacancy, and it is routinely four or five points wider than a marketing package implies.

Step 3: Establish real market rent

Everything downstream of this step depends on getting it right. Loss to lease is the most cited source of value-add upside in multifamily precisely because it requires no capital to capture, which also makes it the easiest number to overstate.

Market rent has to be proven with genuine comparables: similar vintage, similar unit size, similar amenity level, and recently signed leases rather than asking rents. When concessions are in the market, asking rents overstate achieved rents by exactly the amount that matters most.

Then phase the capture over the lease expiration schedule. At a typical 45% to 55% annual turnover, roughly half the gap is addressable per year, so a full mark to market takes eighteen to twenty-four months even when the rent estimate is correct.

Step 4: Size the debt against all three tests

Lenders do not size to one constraint. They test loan-to-value, debt service coverage, and debt yield, then lend to whichever produces the smallest loan.

Which one binds tells you something real about the deal. When DSCR is the constraint, the deal is rate-constrained and will improve if rates fall. When debt yield binds, the income simply does not support the loan at any rate, and only a lower price or higher NOI fixes it.

Size against your underwritten NOI, not the seller's. The lender will.

Step 5: Model the business plan as capital, not as a rent assumption

A renovation program is not a rent increase. It is a capital outlay that produces a rent increase, and the two belong on different lines.

Build the capital budget in three buckets: immediate physical needs, value-add scope, and ongoing reserves. Add 5% to 10% contingency on the renovation scope, because turn costs creep. Then test the whole plan on yield on cost against the prevailing market cap rate.

That spread is the cleanest single measure of whether the plan creates value. Under 100 basis points, the execution risk is usually not being paid for.

Step 6: Pressure-test the exit before you believe the returns

The exit cap rate is applied to the largest cash flow in the model, arriving at the end of the hold, and it is a pure assumption about a future market. It moves returns more than almost anything else you will decide.

The convention that keeps underwriting honest is to expand it 25 to 50 basis points over the going-in rate for a five year hold. Assuming compression is assuming the market does your work for you.

Then run the sensitivity grid. A deal that only clears its hurdle at a compressed exit and above-trend rent growth is a bet on the market, not on the asset. That may still be a bet worth making, but it should be made knowingly.

Step 7: Read IRR and equity multiple together

IRR accounts for timing, which is why committees anchor on it. That same sensitivity means it can be inflated by an early refinance without a dollar of additional profit.

The equity multiple is time-blind and cannot be gamed the same way. Read them together, always:

• High IRR, thin multiple. Check whether a capital event is doing the work.
• Strong multiple, weak IRR. The profit is real but slow. Ask whether the hold period is being paid for.
• Both strong. Now check what share of total profit comes from reversion rather than operations. Above 80% and the deal is an exit-pricing bet.

Where the time actually goes

Done by hand in a spreadsheet, this framework takes most of a day per deal. That is the real constraint on acquisitions volume: not analytical ability, but the hours between an OM landing in your inbox and a defensible number coming out.

Compressing that loop is the entire premise of MultiScreen. Import the OM, T12, and rent roll, and the first six steps run automatically, leaving you the seventh, which is the one that actually requires judgment.

Screen your next deal free →