Office · San Francisco Bay Area, CA

What owners are underwriting in San Francisco Bay Area, CA office

Owners are underwriting a 6.1% going-in cap to a 7.5% exit over a 3-year hold.

Drawn from 18 underwriting models 3E builds and maintains for 1 owners and operators buying office in San Francisco Bay Area, CA.

Updated 2026-10-14

The assumption stack

Exit caps, hold periods, leverage and coverage do not appear in any sale-comparable database. They exist only inside the model a buyer built. These are the medians across 3E's client underwriting.

From 3E client underwriting models18 properties

Exit cap rateThe cap rate the model assumes at sale, which sets the reversion value.

7.5%

7.0%7.5%

middle 50% · 18 models

Exit vs going-in spreadHow much cap-rate softening the underwriting builds in. Computed per deal, then taken as a median.

+140 bps

−82 bps+223 bps

middle 50% · 12 models

Hold periodYears from acquisition to the modelled sale.

3.0 yrs

3.0y4.0y

middle 50% · 18 models

Interest rateThe rate carried in the model at close.

5.50%

5.50%5.50%

middle 50% · 18 models

Operating expense ratioTotal operating expenses as a share of effective gross income.

44%

38%53%

middle 50% · 15 models

Levered IRRThe levered internal rate of return the model projects over the hold.

17.4%

13.2%22.8%

middle 50% · 18 models

A 3-year hold is the convention here. 50% of the 18 models carrying a hold period use exactly that figure. Where a business plan genuinely needs longer, the underwriting is doing something the rest of the market is not — worth being deliberate about.

Pricing

Two independent reads: what 3E's clients are paying, and what the wider market has transacted at. Where they disagree, the gap is usually the story.

Priced into 3E client underwriting18 properties

Going-in cap rateYear-one net operating income divided by purchase price.

6.1%

4.9%7.8%

middle 50% · 12 models

Price per square footPurchase price divided by net rentable area.

$300/SF

$246$414

middle 50% · 18 models

Market sale comparables155 sales · 2022–2026

Going-in cap rateYear-one net operating income divided by purchase price.

6.2%

6.0%9.1%

middle 50% · 14 sales

Price per square footPurchase price divided by net rentable area.

$312/SF

$221$465

middle 50% · 155 sales

Market context

Public reference data for the market, and the hazard profile of the properties 3E is actually underwriting in it.

Metro demographics

Population: 4,653,593
Population, 5-year change: −0.4%
Median household income: $133,780
Median gross rent: $2,426/mo
Renter-occupied share: 45%

US Census Bureau, 2019–2023 ACS 5-year. San Francisco-Oakland-Fremont, CA (CBSA 41860).

Natural hazard

In a FEMA flood zone: 0%
Median national risk percentile: 98
Earthquake risk, high or above: 100%

Across the 18 properties in this cut that carry a FEMA profile. Flood zone = Special Flood Hazard Area (1% annual chance). Risk percentile from the FEMA National Risk Index.

Rate environment

10-year Treasury: 5.21%
5-year Treasury: 5.00%
SOFR, overnight: 3.89%

Federal Reserve Bank of St. Louis (FRED), as of 2026-10-13. The backdrop the rates above were underwritten against.

Method

Where the underwriting figures come from

3E builds and maintains acquisition and asset-management models for owners and operators. The figures above are medians across those models — 18 distinct properties in this cut. They describe what buyers are assuming, not what any property is worth.

Market context

Demographics are US Census Bureau ACS 5-year estimates for the metropolitan statistical area; benchmark rates are published by the Federal Reserve Bank of St. Louis. The hazard profile is not a statistic about the metro — it is computed across the 3E-modelled properties in this cut, using FEMA flood maps and the National Risk Index, and so describes what buyers here are actually choosing to underwrite.

What the ranges mean

The large figure is the median. The two smaller figures are the 25th and 75th percentiles — the middle half of the sample — and the tick marks where the median falls between them. A median hard against one end indicates a skewed sample, which is worth knowing before leaning on the number.

Run this on your deal

Send one OM or T-12 and 3E will underwrite it properly — model, sensitivities, debt sizing and a risk score — benchmarked against the figures above and delivered in the portal within three business days. No charge, and no meeting needed.

Email ebergin@3emanagement.org