Every multifamily underwriting tool demos well on a clean rent roll. Clean rent rolls are not what arrives.
What arrives is an export with merged header cells, unit types labelled three different ways, concessions in a free-text notes column, several units mid-turn, and a total that does not quite tie to the T12. The tool's behaviour on that document is the only thing worth testing, because that is the document you will use it on.
Four specific places the number breaks.
1. Loss To Lease, Read Backwards
Loss to lease is the gap between market rent and in-place rent. It is also two different stories wearing the same number.
A large gap can mean real upside as leases roll. It can equally mean whoever set market rent was optimistic, in which case the gap is not upside — it is a forecast presented as a fact.
The offering memorandum will present the first reading. The test is whether the tool checks market rent against actual recent leases at the property and genuinely comparable ones nearby, and shows you that evidence, or whether it accepts the stated market rent and computes from there.
Ask it: what is market rent based on, and which leases support it.
2. Concessions Not Normalised
One month free on a twelve-month lease is approximately an eight per cent rent reduction. It frequently appears as a one-time line item rather than inside the rent.
Carry the face rent into the model and treat the concession as non-recurring, and income is overstated — then multiplied, because value is a function of that income. On a property with broad concessions the error is not marginal.
The test: put a rent roll with concessions in a notes column in front of the tool and see whether the effective rent comes out reduced, or whether the face rent survives into the model with the concession parked somewhere harmless.
3. Unit Mix Rolled Up Into A Blend
A blended average rent hides the mix that produces it.
A property whose one-bedrooms are full at strong rents and whose three-bedrooms sit vacant can show a healthy blended number and a broken income story. So can one where a renovated tier is carrying an unrenovated one.
A tool should show rent, occupancy and lease expiry by unit type, with the counts. If the output is a single blended figure, you are being shown a conclusion whose composition you cannot check — and the composition is what determines whether the upside is real.
4. Expenses Carried From Somewhere Else
Expense assumptions are where borrowed benchmarks do the most damage, because they are the least visible.
Property taxes reassess on sale in many jurisdictions and not in others — a model carrying the seller's tax line into a post-sale year can be wrong by a material amount before anything else is considered. Insurance has moved sharply and unevenly by geography. Payroll and management differ by property size and operating model.
The test: ask where each expense line came from. "Market standard" is not an answer. The correct answers are the property's own T12, a named comparable, or a stated assumption you can argue with.
The Test That Replaces A Feature Checklist
Take one property you underwrote yourself, with its real documents, and run it. Then compare not the conclusion but the working:
- Is every figure traceable to the rent roll cell, T12 line or comparable it came from?
- Does the tool flag the units it could not parse, or silently drop them? A silent drop is the worst behaviour available, because the total still looks plausible.
- What does it do where evidence is thin — state a range and say why, or produce a confident point?
- Can you change one assumption and have everything dependent move consistently?
Where Titleman Sits
We read the rent roll and the financials against the property record rather than in isolation: figures keep a reference to the document and cell they came from, unit types are rolled up with their counts visible, concessions are normalised into effective rent, and rows that could not be parsed are reported rather than dropped.
Where the comparable evidence is thin, the output says so instead of producing a number that reads confident.
What It Does Not Replace
- A licensed appraisal where a lender or a standard requires one.
- Legal and physical diligence, including the condition of the units behind the rent roll.
- Estoppels, and the tenants' own confirmation of what the leases say.
- The judgement of whoever prices the deal.