Chapter 11: Small multifamily and the art of scaling
A door inside Sofia's $650,000 fourplex costs $162,500. The same door under its own roof, Marcus and Tina's single-family rental, costs $240,000. Both doors hold one tenant, one lease, and one set of appliances, but they sit in different businesses. This chapter covers the 2–4 unit property, the documents that tell the truth about it, the BRRRR strategy with its bruises showing, and the systems that decide whether scaling feels like a portfolio or a second job.
One roof, four rents
Small multifamily, meaning 2–4 units, is the sweet spot where residential financing still applies (chapter 6) but the economics start behaving like a business. The advantages all come from sharing. Four rents share one roof, one foundation, one yard, one insurance policy, one property tax bill, and one address for your manager to drive to.
The sharing shows up in three places. First, the cost of entry per door drops: $162,500 against $240,000 in our pair. Second, the big-ticket items divide. A $20,000 roof on the fourplex is a $200 per-door annual reserve over its 25-year life; the $15,000 roof on the single-family house is $600 a year all by itself. Third, and most important, vacancy stops being binary. When Marcus and Tina's tenant leaves, the property earns $0 and the mortgage continues. When one of Sofia's tenants leaves, the building still collects 75% of its rent.
| Per door | Single-family ($240,000) | Fourplex ($650,000 ÷ 4) |
|---|---|---|
| Purchase price | $240,000 | $162,500 |
| Monthly rent | $1,950 | $1,450 |
| Rent per dollar of price, monthly | 0.81% | 0.89% |
| Roof reserve per year | $600 | $200 |
| Insurance per year | $1,600 | $725 |
| Property tax per year | $3,000 | $1,800 |
| Income lost while one unit sits empty | 100% | 25% |
The fourplex door rents for less in absolute dollars, yet produces more rent per dollar invested and carries thinner fixed costs. That is the whole argument for multifamily in two rows. Chapter 4 already ran Sofia's building through the full NOI, cap rate, and DSCR framework, and chapter 6 traced her DSCR loan, so the underwriting machinery is the same one you already own. What changes at 2–4 units is what you must read before you underwrite anything.
Underwrite the documents, not the listing
A single-family rental gets underwritten from market data. An occupied multifamily gets underwritten from its paperwork, because you are buying the leases as much as the walls. Three documents matter, and the listing is not one of them.
The rent roll is the unit-by-unit list of tenants, current rents, lease start and end dates, and deposits held. The T12 (trailing twelve months) is the actual income and expense history for the past year, real dollars that hit a real bank account. Compare both against the listing's pro forma and you will usually find the pro forma describes a building that does not exist yet: market rents nobody is paying, expenses nobody has achieved. Chapter 5 taught you to rebuild a pro forma; here the T12 does the rebuilding for you. When the T12 and the brochure disagree, the T12 is the building.
The third document protects you from handshake history. An estoppel letter is a short statement each tenant signs confirming their rent, deposit, lease term, and any side agreements with the current owner. Estoppels are how you discover, before closing rather than after, that unit 3 has a verbal deal to pay $200 less for doing the mowing, or that the deposit the seller listed was never actually collected. Make estoppels a condition in your offer on any occupied building.
Inherited tenants and the humane value-add
Multifamily listings love the phrase "rents below market," and for once the listing is often right. Long-time owners drift; a tenant of eight years quietly ends up paying 2019 rent. Buying that building is the classic value-add play: bring rents to market and the same walls produce more NOI.
The honest version respects two costs. The turnover cost is arithmetic: pushing a $300 raise on day one can produce a move-out that costs a month of vacancy, a $1,200 turn, and a placement fee, roughly $3,400, most of the first year's gain. The human cost is not arithmetic, and it is real: an inherited tenant is a person who kept the building you just bought alive. The workable path is gradual, lawful, and boring. Follow your state's notice rules, raise in steps, and take units to full market rent at natural turnover.
Sofia's fourplex closed with three units at or near the $1,450 market and one long-time tenant at $1,150. The seller's broker called the gap "$3,600 a year of instant upside." Sofia called it a person named Gloria with twelve years of on-time payments. She raised the rent $150 with sixty days' notice, then another $150 a year later, and told Gloria the plan in writing up front. Eighteen months in, the unit sits at market, Gloria stayed, and Sofia never paid for a turnover. The spreadsheet got its $3,600; it just had to wait a year and a half for it.
BRRRR, in full, bruises included
Chapter 7 previewed BRRRR: buy, rehab, rent, refinance, repeat. The idea is to buy a distressed property cheaply, renovate it, rent it, then take a cash-out refinance against the new, higher value to recover your capital for the next deal. Social media presents it as an infinite money loop. It is actually a financing strategy stapled to a construction project, and construction projects have opinions of their own. Sofia's most recent one is worth tracing line by line, plan against reality.
| Step | The plan | What happened |
|---|---|---|
| Purchase | $95,000 | $95,000 |
| Closing costs | $3,500 | $3,500 |
| Rehab | $35,000 | $42,000 (20% overrun) |
| Holding costs | $4,500 (6 months) | $6,000 (8 months) |
| All-in cost | $138,000 | $146,500 |
| Appraisal | $195,000 hoped | $185,000 (about 5% short) |
| Cash-out refinance | 75% LTV: $146,250 | 70% LTV: $129,500 |
| Refinance closing costs | −$4,000 | −$4,000 |
| Cash back out | $142,250 | $125,500 |
| Cash left in the deal | $0, plus $4,250 extra | $21,000 |
The plan column is the version in every webinar: all of her money back and change. Reality hit both of BRRRR's classic failure points. A foundation surprise and a plumbing re-route pushed the rehab from $35,000 to $42,000, and the extra two months added $1,500 of holding costs. Then the appraiser, unmoved by her renovation budget, came in at $185,000 instead of the hoped-for $195,000. Appraisals price the comparable sales, not your receipts.
At 75% LTV the new loan would have been $138,750 at 7.9%, a $1,008 monthly payment, and the stabilized numbers go slightly negative: minus $13 a month. So Sofia borrowed less on purpose. At 70% LTV the loan is $129,500, the payment is $941, and the house cash flows:
| Stabilized year, BRRRR house | Amount |
|---|---|
| Rent ($1,750 × 12) | $21,000 |
| Vacancy (7%) | −$1,470 |
| Property tax | −$2,400 |
| Insurance | −$1,100 |
| Management (8% of collected) | −$1,562 |
| Maintenance (6%) | −$1,260 |
| Capex reserve (6%) | −$1,260 |
| NOI | $11,948 |
| Debt service ($129,500 at 7.9%) | −$11,292 |
| Cash flow | $656 (about $55/mo) |
The honest ending: Sofia turned $146,500 of cash into a renovated house worth $185,000 with $55,500 of equity, a thin positive cash flow, and $125,500 recycled for the next project, while $21,000 of her money stays stuck in the walls. That is a good outcome. It is capital recycled, not capital multiplied, and the return on the $21,000 left behind is ordinary, not infinite. Two practical notes sharpen the picture: most lenders want 6–12 months of seasoning (ownership history) before they will lend against the appraised value rather than your purchase price, and anyone using hard money instead of cash pays points and double-digit interest while the rehab runs over.
Underwrite every BRRRR with the rehab 20% over budget and the appraisal 5% short. If the deal still works, proceed. If it only works at the contractor's first bid, it does not work.
Systems: making three doors feel like one
Scaling fails in the bookkeeping before it fails in the buildings. The owners who last set up the boring infrastructure early.
Set these up before door three, not after door five
- One bank account per property. Every rent in, every expense out, one place. Your Schedule E and your sanity both depend on it.
- Software for rent collection, maintenance tickets, and lease storage, even at two doors. Venmo and a shoebox do not scale.
- A maintenance bench found before the emergency: a handyman, a plumber, an electrician, and an HVAC tech who answer your calls because you pay promptly.
- A reserve per property (chapter 3's 3–6 months of PITIA), not one shared pool that a single roof can drain.
- A standing rhythm: rents reviewed yearly, insurance re-shopped every two years, and a walk-through of every unit annually.
When to stop counting doors
Door count is the vanity metric of real estate. Ten doors that net $250 each while eating your weekends are a worse business than four doors that net $400 each and run through a manager. The score that matters is cash flow per hour of your time. If your portfolio pays $2,400 a month and takes ten hours, you earn $60 an hour as an investor. If the next two doors add $300 a month and eight hours, those doors pay $9 an hour, and you would do better tutoring. Sofia caps her growth not at a door count but at a rule: every new building must keep her total time under one hour per door per month, or it must budget for management that does.
Doors are not points. Measure cash flow per hour of your time, and let that number, not the door count, decide whether you buy the next building.
Where people go wrong
- Underwriting the brochure. The rent roll and T12 describe the building you are buying; the pro forma describes the seller's dream.
- Skipping estoppel letters. Side deals and missing deposits surface at closing or in court; a one-page letter surfaces them for free.
- Raising every inherited rent on day one. The turnover math usually eats the first year's gain, and the building's reputation eats the second's.
- Running BRRRR on the contractor's first bid. Budget the overrun and the short appraisal; be pleasantly surprised when only one shows up.
- Scaling the doors but not the systems. Mixed bank accounts and a missing maintenance bench turn door five into a part-time job with no salary.
Key takeaways
- Small multifamily wins on shared costs: a door in Sofia's fourplex costs $162,500 and rents at 0.89% of price monthly versus 0.81% for the single-family door, and one vacancy costs 25% of income, not 100%.
- Underwrite occupied buildings from the rent roll and T12, and require estoppel letters; the listing pro forma is marketing.
- Inherited below-market rents are real upside best captured gradually and lawfully; day-one raises usually trade the gain for a turnover bill.
- BRRRR recycles capital but rarely all of it: with a 20% rehab overrun and a 5% short appraisal, Sofia recovered $125,500 of $146,500 and left $21,000 in the deal, on purpose, to keep the cash flow positive.
- Systems (an account per property, software, a maintenance bench) make scale survivable, and cash flow per hour, not door count, is the score.
Sources: Fannie Mae Eligibility Matrix · IRS Publication 527: Residential Rental Property