Most people think multifamily underwriting starts when the spreadsheet opens.
It doesn't.
It starts when you ask a much less comfortable question:
Can I trust these numbers enough to make a decision with real money behind it?
That's the real start.
A broker OM can look polished and still be soft. A T12 can look clean and still bury weak expenses. A rent roll can show strong occupancy and still hide a revenue problem. And once you get into larger multifamily deals, especially 50+ units and really 100+ units, those little misses don't stay little. They stack. Then they spread.
That's underwriting.
Not spreadsheet theater. Not vocabulary practice. Not trying to "make the deal work."
It's the process of taking the story apart, line by line, until you understand what's real, what's shaky, and what breaks first if the plan doesn't go perfectly.
What multifamily underwriting actually is
In plain English, multifamily underwriting is the process of deciding whether an apartment deal makes financial sense.
That's the short version.
The more honest version: underwriting is the process of taking a messy set of property documents, market assumptions, and financing terms, and turning all of it into one decision:
Is this deal worth pursuing at this price, with this risk, under these conditions?
That's the job.
I'm trying to answer a few basic questions:
- What does this property actually earn?
- What does it really cost to operate?
- What happens once debt gets layered on?
- Which assumptions are grounded, and which ones are doing too much heavy lifting?
- If the plan gets stressed, how badly does the deal bend?
Lenders do it. Buyers do it. Equity partners do it. Acquisition teams do it. They may emphasize different things, but the core question is the same:
Does the story still hold up when the numbers get honest?
What multifamily underwriting is not
This section matters because a lot of bad habits start right here.
Multifamily underwriting is not:
- copying broker numbers into a model
- calculating a cap rate and calling it analysis
- trusting a tidy T12 because it looks professional
- assuming physical occupancy means healthy revenue
- deciding a deal is good because the year-one return looks decent
- building a model that "works" only because the assumptions were never challenged
A pretty spreadsheet can still hold bad judgment.
That's one of the more expensive lessons in this business.
I've seen deals where the summary page looked sharp, the returns looked attractive, and the whole thing started to wobble the minute you pushed on concessions, bad debt, taxes, insurance, or actual lease-level rents. It doesn't take fraud to break a deal. Most of the time, it's a stack of smaller assumptions that were just a little too generous.
Why underwriting matters more on larger multifamily deals
If you are underwriting a duplex or a fourplex, people sometimes get away with a looser process.
Not a good idea. But it happens.
Once you move into larger residential multifamily, that margin disappears fast.
Why?
Because scale magnifies sloppiness.
On a 180-unit deal:
- a weak insurance assumption can hit value harder than people expect
- a few delinquent units can distort economic occupancy
- a soft concession trend can quietly drag effective rent down
- a handful of down units can create a very different revenue story than "95% occupied" suggests
- a 25-basis-point move in exit cap can matter more than a year of optimistic rent growth
That's the point.
Larger multifamily underwriting is not just smaller underwriting with more units. It is a different level of discipline.
Underwriting workflow
Multifamily underwriting starts with raw files, but the real work is turning those files into a decision you can actually defend.
The core documents you need
If you are missing these, your underwrite is already weaker than it should be.
Required document checklist
- Rent roll
- T12 operating statement
- Trailing collections / delinquency detail
- Debt quote or financing assumptions
- Real estate tax information
- Insurance information or current quote guidance
- Capex scope / renovation plan
- Rent comps
- Sales comps
- Market / submarket context
- OM or broker package
You do not always get every item at the start. That's normal.
But the further you get into a deal without this file stack tightening up, the more assumptions you are making in the dark.
The core inputs in a multifamily underwrite
1) The rent roll
This is the unit-by-unit lease snapshot. It shows who is in place, what they are paying, what is vacant, whether future residents are showing up, whether balances are building, and whether the in-place rent story matches what people are claiming.
A rent roll is one of the fastest reality checks in the whole process.
If you don't know how to read one properly, that's one of the first gaps to fix.
2) The T12
The trailing 12-month operating statement shows how the property has actually performed over the last year. Income. Expenses. Monthly movement. Sometimes noise. Sometimes trouble.
It is important. It is useful. It is not self-explaining.
A T12 tells you what happened recently. It does not tell you what happens next.
3) Collections and delinquency detail
This one gets ignored too casually.
A property can look healthy on the rent roll and still have a soft revenue engine if collections are weak or balances are building.
4) Debt terms
Interest rate. Amortization. Leverage. IO period. DSCR limits. Debt yield limits. Refinance logic.
Some deals look strong right up until the debt gets honest.
Then they stop looking strong.
5) Taxes and insurance
These two lines deserve more attention than they usually get.
They are large enough to move value in a real way, and they are two of the easiest places for less experienced underwriters to stay too optimistic.
6) Capex plan
If the business plan depends on renovating units, fixing deferred maintenance, repositioning the asset, or upgrading amenities, that has to be reflected clearly.
Capex is not a footnote. It is part of the economics.
7) Rent comps and sales comps
Rent comps test the revenue story. Sales comps help frame pricing.
Useful? Absolutely.
Reliable without judgment? Not even close.
8) Market and submarket context
You need to understand supply pressure, rent trends, employer base, affordability, demand drivers, and neighborhood risk.
Underwriting without market context is spreadsheet math pretending to be analysis.
The core metrics
These are the numbers people love to throw around in multifamily.
Some of them matter a lot. Some get abused. A few do both.
If you're newer to underwriting, here's the trap: you start hearing terms like cap rate, DSCR, IRR, and debt yield so often that they start sounding more useful than they actually are on their own. They are useful. Just not in isolation.
Effective Gross Income (EGI)
EGI is what the property actually keeps after vacancy, concessions, and bad debt, plus other income.
That's a much better revenue number than scheduled rent.
This is one of the first places newer investors get fooled. They see the rent roll, multiply the scheduled rents, and feel like they understand the revenue story. Not really. Not until they know what is leaking out of it.
Net Operating Income (NOI)
NOI is EGI minus operating expenses, before debt service and below-the-line items.
This is the number that keeps showing up for a reason.
NOI drives value. It shapes debt sizing. It tells you how much real operating cushion the property has before financing gets involved. If your NOI assumption is soft, the rest of the model can look polished and still be wrong.
Cap Rate
Cap rate is a pricing yield, usually framed as NOI divided by purchase price or value.
Useful? Absolutely.
Overused? Also yes.
Cap rate is a quick way to compare pricing across deals. It is not some magic truth serum. Two deals can have the same cap rate and very different risk.
DSCR
Debt Service Coverage Ratio measures how comfortably NOI covers debt service.
Lenders care about this a lot. They should.
If DSCR is thin, the deal has less room to absorb mistakes. That does not automatically kill it, but it should make you pay attention.
Debt Yield
Debt yield measures NOI relative to the loan amount.
This one does not get talked about as much outside more serious underwriting circles, but it matters. It is one of the cleaner ways to look at lender risk because it is less distorted by amortization and rate structure than some other metrics.
Break-Even Occupancy
This tells you how much occupancy the property needs to cover operating expenses and debt service.
Not the sexiest metric in the model. Still one of the more revealing ones.
A fragile deal usually starts looking fragile here.
Cash-on-Cash Return
This shows the annual cash yield on invested equity.
Helpful if you care about current income. Less helpful if you use it like a shortcut for total return quality.
A deal can show decent cash-on-cash and still be carrying a weak exit story or too much operational risk.
IRR
Internal Rate of Return measures the timing-adjusted return on invested equity over the hold period.
Important metric. No question.
It is also one of the easiest metrics to dress up. A slightly aggressive exit cap, a smoother-than-realistic renovation pace, or a little too much confidence in rent growth can make the IRR look a lot prettier than it deserves.
Equity Multiple
Equity multiple measures total cash distributions relative to total equity invested.
It is simple, which is part of why people like it.
But simple cuts both ways. It does not care whether you doubled your money in three years or ten. That matters.
Quick-reference metrics table
| Metric | What it measures | What people often miss |
|---|---|---|
| EGI | Real revenue after leakage | Scheduled rent is not collected rent |
| NOI | Income after operating expenses, before debt | NOI is not investor cash flow |
| Cap Rate | Pricing yield on NOI | Similar cap rates can hide very different risk |
| DSCR | NOI coverage of debt service | A "passable" DSCR does not make the deal safe |
| Debt Yield | NOI relative to loan amount | It can cap proceeds faster than expected |
| Break-Even Occupancy | Occupancy needed to cover expenses and debt | It shows how little room for error a deal may have |
| Cash-on-Cash | Current annual cash yield on equity | It says very little about the full hold-period story |
| IRR | Timing-adjusted return | Exit assumptions can flatter it fast |
| Equity Multiple | Total return relative to equity invested | It ignores time |
A quick example of what underwriting is really catching
Let me make this more real.
Say you are looking at a 120-unit deal.
The broker says the property is 95% occupied. Sounds solid. The market-rent story looks reasonable. The T12 looks stable enough. At first pass, the deal feels fine.
Then I slow down.
- The rent roll shows a cluster of discounted leases.
- A few residents are carrying balances.
- The T12 shows credit loss drifting the wrong way.
- Insurance was underwritten off the trailing figure instead of a fresh quote.
- Taxes were kept too close to history even though the acquisition basis suggests reassessment pressure.
Now the deal feels different.
Not broken, necessarily. But different.
That's underwriting.
It is not always about catching dramatic fraud. More often, it is about catching the stack of smaller assumptions that were quietly flattering the story.
The first 10 minutes: what I verify first
When I want a quick first-pass read on a multifamily deal, this is where I start:
- Does the rent roll support the occupancy and rent story?
- Are concessions heavier than the headline numbers suggest?
- Is bad debt or delinquency showing signs of stress?
- Do taxes and insurance look too close to trailing history?
- Are there down units or offline units being treated like ordinary vacancy?
- Does the business plan depend on rent growth that the comps do not really support?
- If exit cap moves slightly, does the deal still make sense?
This is not the full underwrite.
It is the first filter. And it catches more than people think.
The real underwriting process, step by step
This is the part a lot of articles oversimplify.
They make underwriting sound like a clean sequence where you move from input to output and the answer politely reveals itself. Real life is messier than that. Files are incomplete. Sellers categorize things oddly. Brokers frame things in the best light possible. Debt markets move. Insurance quotes come in ugly. Taxes do what they do.
Still, the process itself is straightforward.
- Review the raw files. Start with the rent roll, T12, collections, OM, debt assumptions, and comps. And actually read them. Not just the totals. Not just the summary tab. Read the file set closely enough to figure out where the weak spots probably are before you start modeling. That alone will put you ahead of a lot of people.
- Normalize income. Now start cleaning up the revenue story. This usually means asking: What income is recurring? What is noisy? What is inflated? What is missing? What is technically there but not durable? Vacancy, concessions, bad debt, loss to lease, other income, down units, future residents, charge-code quirks. This is where the deal starts becoming real.
- Normalize expenses. This is where people either get disciplined or get lazy. Some historical expenses are too low because spending was deferred. Some are too high because of a one-time issue. Some are categorized in ways that make sense for bookkeeping but not for underwriting.
- Build the debt assumptions. Now let the financing do its job and put pressure on the deal. Rate, amortization, IO period, leverage, refi assumptions, DSCR limits, debt-yield limits. A lot of deals still look great before this step. Fewer do after.
- Model the business plan. This is where optimism usually tries to sneak back in. Maybe the plan is to renovate units, push rents, tighten collections, clean up expenses, and improve operations. Fine. But every one of those lines needs support. Not energy. Support. If the plan depends on premiums the comp set does not really support, or on a renovation pace that sounds great in a slide deck but not in real operations, that matters.
- Run sensitivity cases. This is where underwriting stops being a story and starts becoming a decision tool. Push on the assumptions. What happens if exit cap moves out a bit? What happens if insurance lands materially higher? What happens if tax reassessment bites harder than expected? What happens if concessions linger? What happens if rent growth is just okay instead of amazing? A deal that only works in one neat version of the future is telling you something.
- Decide whether the risk is worth it. This is the point. Not whether you can force the model to show acceptable returns. Usually, you can. The real question is whether the risk-adjusted opportunity is strong enough to deserve your time, your team's attention, and your capital stack's confidence. That's the standard.
Process recap
A clean snapshot of the seven steps and the question each one is really trying to answer.
| Step | What happens | What you are really trying to answer |
|---|---|---|
| 1. Review the files | Read rent roll, T12, collections, OM, debt terms | Do the raw documents look trustworthy? |
| 2. Normalize revenue | Adjust for vacancy, concessions, bad debt, other income quality | What is real revenue? |
| 3. Normalize expenses | Recast historical expenses into a believable operating load | What should it really cost to run? |
| 4. Build debt assumptions | Size and stress debt using realistic financing terms | What does the capital stack do to the deal? |
| 5. Model the business plan | Test renovation, rent-growth, and operational assumptions | Does the plan actually hold up? |
| 6. Run sensitivities | Stress exit cap, rents, expenses, taxes, insurance, timing | Where does the deal get fragile? |
| 7. Make the decision | Compare risk, reward, durability, and margin of error | Is this worth pursuing at this basis? |
What changes on 50+ and 100+ unit deals
Scale changes the whole feel of the analysis.
Concessions matter more
A single discounted lease on a small property may not move much. On a larger asset, concession strategy can materially reshape effective revenue.
Payroll matters more
Bigger assets need real staffing logic. Leasing, maintenance, turns, collections, resident issues. Those costs are not theoretical.
Collections quality matters more
A property can look full and still have a weak revenue engine if enough residents are carrying balances or if effective rent is softer than the headline numbers suggest.
Down units matter more
You need to separate vacant-but-rentable from vacant-and-offline. People blur those together too casually.
Taxes and insurance matter more
The larger the asset, the larger the absolute swing from weak assumptions here.
Operations matter more than theory
This is where some investors get humbled.
A model can look smart. The property still has to function in the real world. On larger assets, that operational reality shows up fast.
Small-property mindset vs larger multifamily reality
| Topic | Smaller property mindset | Larger multifamily reality |
|---|---|---|
| Vacancy | One or two units can distort the picture | Vacancy becomes a broader revenue-management issue |
| Payroll | Often minimal or lightly outsourced | Staffing model matters materially |
| Concessions | Sometimes minor | Can reshape effective revenue |
| Delinquency | Easier to notice manually | Can hide inside a larger file set |
| Insurance | Important but smaller absolute swing | Can materially hit NOI and value |
| Turns / down units | More visible one by one | Need systematic review |
| Debt sizing | Often simpler | DSCR and debt yield discipline matter more |
Red flags I would verify first
If I am reviewing a deal quickly, these are some of the first things I want to verify before I trust the story too much:
That is not the full underwrite.
It is the first pressure test.
Common underwriting mistakes
There are plenty. These are the ones I see the most, especially when someone is still early in building real underwriting discipline.
Trusting the OM too quickly
The OM is useful. I'm not knocking it.
But it is still marketing.
That means it can be directionally helpful and still financially flattering at the same time.
Ignoring concessions because "occupancy looks good"
This one gets people more than it should.
A property can be 95% occupied and still have a weaker revenue story than it first appears. Concessions, discounts, credits, and soft effective rent tend to hide under nice-looking occupancy numbers.
Mixing capex and opex
This is one of the easiest ways to make a deal look cleaner than it is.
Sometimes it happens innocently. Sometimes it is just lazy categorization. Either way, if you do not separate recurring operating burden from capital work correctly, the model gets softer than it should be.
Underwriting taxes too close to trailing history
This is a classic mistake, especially after a sale at a meaningfully higher basis.
Historical taxes are useful. They are not always predictive.
Underwriting insurance off a stale number
Same problem.
If the market says insurance is going to hit harder than trailing performance suggests, the trailing number is not conservative just because it is historical.
Assuming market rent is automatically real
I see this one all the time.
People grab the "market rent" field, plug it into the business plan, and move on. But the market-rent story still has to survive the comp set, the current lease trade-out, the property condition, and the actual submarket context.
Getting too optimistic on exit cap
Somewhere, quietly, this is making a lot of IRRs look smarter than they are.
A little optimism here goes a long way. Too long a way.
Ignoring delinquency, future tenants, and down units
These are the details that people wave off because they feel small.
On a larger multifamily deal, they are not small.
They are exactly the kind of things that tilt the story while everyone is still staring at the headline metrics.
How institutional investors think differently
This is not about sounding sophisticated. It's about having better habits.
Institutional-style underwriting usually does a few things better:
- assumptions are documented, not implied
- historical reality is separated from forward assumptions
- downside gets tested, not just upside
- process is repeatable
- durability matters more than presentation
Institutional investors do not get paid for optimism. They get paid for disciplined decision-making.
That mindset is worth borrowing even if you are not an institutional shop.
Actually, especially if you are not.
Because smaller teams usually have less room to absorb bad mistakes.
Where technology fits
Technology should make underwriting faster and cleaner.
Not lazier.
That distinction matters.
MSA did not come out of a brainstorm session. It came out of real underwriting friction I kept running into at XSITE Capital, where I serve as an Underwriting Specialist. XSITE currently reports a $276 MM portfolio value and 1,422 doors under management, so the pain points were real: messy rent rolls, inconsistent T12s, time lost in cleanup, and the constant pressure to get to a defensible decision faster.
This is also the friction that pushed me to build MSA.
Specifically, four tools handle the parts that drag underwriting down:
- QuicRollAI helps clean and structure messy rent roll and T12 files.
- MSA Direct helps move parsed data into the underwriting workflow with less manual re-entry.
- MSA Analyzer helps turn the cleaned data into a full underwriting model.
- MSA IQ provides submarket data to aid in underwriting and help investors better understand the market around the deal.
That is where technology belongs.
It should reduce the time spent fighting files, so more time can go toward judgment, assumptions, and decision-making.
Still, judgment stays human. It should.
Fast workflows are helpful. Fake certainty is not.
If you want to go deeper on the two files that drive a huge share of the underwriting story, the next places to spend time are the T12 and the rent roll.
Final takeaway
Here's the cleanest way I know to say it:
Multifamily underwriting is the process of turning messy documents, market assumptions, and debt terms into a real investment decision.
That's it.
Not a spreadsheet contest. Not a vocabulary test. Not a sales exercise.
A decision.
And the best underwriting usually has the same habit built into it: it spends less time trying to prove the deal works and more time figuring out where it breaks.
That's the habit worth building.
The best underwriting spends less time trying to prove the deal works and more time figuring out where it breaks.
FAQ
What is multifamily underwriting in simple terms?
It is the process of figuring out whether an apartment deal actually makes financial sense once you account for the real income, real expenses, real debt, and real risk.
What documents do you need to underwrite a multifamily deal?
At a minimum, I want the rent roll, T12, collections or delinquency detail, debt assumptions, tax information, insurance guidance, comps, and a capex plan if the business plan depends on upgrades.
What is the difference between underwriting and due diligence?
Underwriting is the upfront decision work. You are trying to decide whether the deal deserves to move forward.
Due diligence happens after that and goes deeper. That is where you verify the documents, physical condition, leases, contracts, legal issues, and all the other things that can still surprise you if you got too comfortable too early.
What metrics matter most in multifamily underwriting?
The main ones are EGI, NOI, cap rate, DSCR, debt yield, break-even occupancy, cash-on-cash return, IRR, and equity multiple.
That said, the "most important" metric usually depends on what part of the deal you are trying to pressure-test.
Can you underwrite a multifamily deal in Excel?
Yes. Plenty of serious investors still do.
Excel is not the weak point. The weak point is usually the file cleanup before Excel, or the assumptions getting fed into it.