Deal type
Apartments, on a unit mix
A rent roll is the wrong shape for a hundred and twenty apartments. You do not underwrite them one lease at a time; you underwrite four unit types, a turnover rate, and whatever the plan is for the units that are under market. That is what this models.
A row is a unit type
A studio, a one-bed, a two-bed, a three-bed. Counts, areas, market rent and what the current leases actually pay. The gap between those last two is loss to lease, and it is on the row rather than buried in a total.
- A broker’s rent roll imports as a mix. Rows are grouped into unit types, and a vacant unit’s stale rent is left out rather than averaged in.
- In-place left blank means market. The assumption that adds nothing, rather than one that invents a discount nobody entered.
- The totals line is the model's. 120 units, 99,160 SF, $2,366,400 of potential rent, $105,720 below market: computed, not captioned.
The value-add plan, as a plan
Renovating forty units at twelve thousand each and pushing rents two hundred is the most common thing a small firm does to an apartment building. It belongs in the model as a schedule, not as a nudged rent number.
- Scope per unit type, pace per building. "Eight a month, all the two-beds and half the one-beds", which is how the plan gets described.
- Capital, below NOI, on its own line. A renovation taken out of NOI would be capitalized into the exit at the cap rate, which is a valuation nobody would defend.
- The premium moves potential rent. A renovated unit's market rent is the higher one; leaving it out would make a finished program look like the building was still under market on rents it is already charging.
Quoted in doors
A broker says a hundred and twenty-three thousand a door. They do not say a hundred and forty-nine a foot. The returns strip carries price per unit when there are units, and a lender sizes against NOI per unit.
- Reserves per door. Two-fifty to three-fifty a unit a year is how a multifamily reserve is quoted and how a lender sizes one.
- RUBS, parking, pet rent priced per unit per month, reading the mix's own count. Change the mix and the billback follows.
Two figures nothing lets you contradict
Physical vacancy is whatever the turnover rate and the downtime imply. Fifty turns a month held empty for three weeks each is that much vacancy, always. And loss to lease burns off over one full turn of the building, so 45% turnover means twenty-seven months.
- Neither is a field. Storing them would let somebody set a vacancy the turnover rate contradicts, and then the screen and the model would describe different buildings.
- Credit loss is separate, on the Assumptions tab, applying to rent actually billed, so the same empty unit is never counted twice.
- A new building leases up to it. Give the delivery month, the share pre-leased and the leases a month, and the building fills in a straight line to the occupancy turnover leaves room for. Free rent and a leasing cost ride on each lease-up lease, and nothing operates before delivery.
Swept on the things that decide it
A value-add deal lives or dies on whether the premium turns out to be what you assumed and whether the work costs what you were told. Both are levers, and the second grid opens on exactly that pair.
- Turnover moves two things at once, deliberately: it sets the physical vacancy and the pace loss to lease burns off. A building that turns faster is emptier and reaches market sooner, and you cannot sweep one without the other because that combination cannot happen.
- The findings panel says when a renovation will not finish before the sale: units priced with their cost spent and none of their premium in the forward NOI the exit capitalizes.
Honest
What it does not do
A page that only lists strengths is a page you have to check somewhere else.
- No per-unit detail. A specific apartment with a lease expiring in March is not expressible. That is the deliberate trade for four rows instead of two hundred, and on this model it will stay that way.
- Loss to lease closes on a straight line over one full turn of the building. Defensible, and checkable on a napkin, which is why it was chosen over something more elaborate that nobody could verify.