What San Francisco’s $440 Million Apartment Sale Actually Tells Investors

Reported purchase price

$440M

59 San Francisco apartment buildings

Reported apartments

928

About 15.7 units per building on average

Price per apartment

$474K

Reported price divided by reported units

Data: Reported portfolio transaction as of November 4, 2026; Rent Board housing-inventory methodology checked November 8, 2026.

A $440 million apartment sale makes an easy headline. The more useful number is $474,138 per unit, and even that number is only the beginning of the underwriting.

Fort Point Real Estate Partners reportedly bought 59 San Francisco residential buildings containing 928 apartments. The portfolio spans neighborhoods from Pacific Heights and Russian Hill to Hayes Valley and the Haight. The Registry reported that the basis was about 19% below the prior owners’ cost per unit.

My read is simple: the sale confirms that large buyers will put serious capital into San Francisco apartments at the right basis. It does not establish a citywide value per unit, prove a market bottom, or tell you whether a particular building will cash flow.

Start with the arithmetic

The $7.46 million average building price is nearly useless without the building mix. A 16-unit property in Nob Hill and a four-unit property in the Haight can share one portfolio but have completely different legal rents, expense loads, seismic work and debt capacity. The 15.7-unit average does not mean the portfolio contained 59 similar 16-unit buildings.

The same problem applies to the reported 19% discount from the prior owners’ cost per unit. The current purchase is a 59-building subset of a larger portfolio that went through a deed in lieu. Without a matched list of properties and their original allocations, that percentage is useful context, not a same-property appreciation index.

What NOI would support a $474,138 basis?

Price per unit becomes more useful when you connect it to income. If one apartment carries $474,138 of purchase price, the annual NOI needed to produce a 3.5%, 4.5% or 5.5% unlevered cap rate is just price multiplied by the target cap rate.

At a 4.5% cap rate, the reported basis needs about $21,336 of annual NOI per unit, or $1,778 a month. That is not the rent target. Gross rent must also cover vacancy, collection loss, property tax, insurance, utilities, repairs, management and the rest of the operating budget before it becomes NOI.

This is why I would not take a citywide rent headline and back into a valuation. A rent-controlled unit with a long tenancy can collect far less than a renovated vacancy in the same building. Two buildings on the same block can also carry different property-tax bases, insurance histories and deferred work.

The real investment thesis is operating income

Fort Point’s earlier company announcement described a strategy of seismic and fire-life-safety work, unit and common-area renovations, and activating underused retail. That is more informative than the portfolio average because it points to where the buyer expects to spend money and create value.

We focus on cash-flowing investments. Better operations, lower vacancy, disciplined expenses and well-chosen capital work can improve recurring NOI. When that cash flow becomes more durable, the asset itself can appreciate. That is a much stronger plan than relying on market appreciation or rent growth above inflation, both of which sit outside the owner’s control.

The recurring mistake in San Francisco multifamily is to count the future twice: first as a higher rent, then again as a lower cap rate. If the acquisition only works when rents jump and the market pays more for each dollar of NOI, the margin of safety is imaginary.

What the portfolio headline leaves out

The SF Rent Board Housing Inventory is useful diligence context, but it is owner-submitted and changes daily. The city says each row is a housing unit and warns analysts to count unique IDs within a selected reporting year rather than summing rows. It is a record to reconcile against leases and property files, not a substitute for them.

How I would use this sale on a smaller deal

1. Calculate the subject property’s price per legal unit, but do not stop there.
2. Build current NOI from documents, not market rent or a broker’s blended pro forma.
3. Divide current NOI by price and compare the resulting cap rate with truly similar closed sales.
4. Add a separate five-year capital schedule and cash reserve. Do not hide major work inside an expense ratio.
5. Run debt service using a property-specific quote and a higher-rate downside case.
6. Show renovation, vacancy and rent-reset opportunities separately from the base case.
7. Decide whether the building works with flat real rents. Treat market appreciation as upside.

Our investment-property underwriting guide explains the full workflow. The San Francisco cap-rate calculator lets you test the actual rent roll, vacancy, expenses and debt, and the multifamily loan guide explains why the financing line changes at five units.

My conclusion

The $440 million sale is a strong vote for San Francisco apartment ownership, but it is not a shortcut to valuation. The headline basis tells you where to start asking questions. The rent roll, capital plan and debt tell you whether the investment works.

I would rather buy one building with defensible current cash flow than 59 buildings that need the market to rescue the model.