San Francisco Rental Yield: 3.98% Gross, 1.81% Net, and the Rent-Control Problem
San Francisco's rental yield worked from published data, plus the thing no yield figure for this city can ignore — you buy the tenancy, not the asking rent.
On this page
San Francisco's ratio of asking rent to mid-tier home value is 3.98% as of September 2026 — rent $4,784 a month against a value of $1,441,021. That ranks 30th of the 92 Bay Area cities in this dataset — a little above the regional median of 3.40%, and behind Oakland at 4.66%, East Palo Alto at 5.14% and South San Francisco at 4.09%.
It is also the figure most likely to mislead, for a reason specific to this city.
The arithmetic first
| Line | Amount / year | Basis |
|---|---|---|
| Gross rent | $57,408 | $4,784/mo × 12 |
| Property tax | $15,851 | 1.1% of $1,441,021 |
| Insurance | $2,200 | estimated |
| Vacancy | $2,870 | 5% of rent, assumed |
| Maintenance | $5,741 | 10% of rent, assumed |
| Management | $4,593 | 8% of rent, assumed |
| Total costs | $31,255 | 54.4% of gross rent |
| Net operating income | $26,153 | |
| Net yield | 1.81% |
With 25% down ($360,255) at 7% over 30 years, debt service is $86,284 a year against $26,153 of net operating income — negative $60,131 a year, a cash-on-cash return of −15.5% on $389,075 of cash in.
The part that makes San Francisco different
$4,784 is an asking rent. You do not buy asking rents; you buy a building with people in it.
San Francisco's rent stabilisation ordinance limits annual increases on covered units — broadly, multi-unit buildings with a certificate of occupancy before 13 June 1979 — to a percentage set each year by the Rent Board. Over a long tenancy, the gap between the stabilised rent and the market asking rent compounds into something very large.
Three consequences for a yield calculation, none of which appear in the ratio above:
1. A tenanted building's actual rent roll may be far below the index. If a covered unit's sitting rent is 40% below market, the real gross yield on that unit is 40% below 3.98% — about 2.4%, not 3.98%. The asking-rent index describes the unit you could rent if it were empty and uncovered, which is precisely the unit that is not for sale.
2. The upside is a vacancy event, not a rent review. Under California's Costa-Hawkins Rental Housing Act, a landlord may reset to market rent when a covered unit becomes vacant voluntarily. So the value of a below-market tenancy is an option that pays out on an unpredictable date, and underwriting it as income is underwriting a guess about when someone will move.
3. Buildings built after June 1979 are not covered, and single-family homes and condos are largely exempt from price controls under Costa-Hawkins. A newer condo therefore gets much closer to the index figure than a 1920s four-unit building does — and the two are both "San Francisco rental property".
Statewide rules also apply
California's Tenant Protection Act of 2019 (AB 1482) caps annual increases at 5% plus local CPI, to a maximum of 10%, on most residential property more than 15 years old statewide — including in cities with no local ordinance. It also requires just cause for eviction after 12 months. Single-family homes and condos are exempt when the owner is not a corporation and the tenant has been given the required written notice.
So "no local rent control" does not mean "no cap". For a yield model, the practical floor across almost all of the Bay Area is that you cannot assume rents reset to market annually.
How San Francisco compares
| City | Gross | Net | Costs as % of rent |
|---|---|---|---|
| Oakland | 4.66% | 2.18% | 53.3% |
| San Francisco | 3.98% | 1.81% | 54.4% |
| Hayward | 3.90% | 1.63% | 58.1% |
| San Jose | 3.17% | 1.18% | 62.8% |
| San Mateo | 2.84% | 0.96% | 66.3% |
San Francisco sits in the middle: better than most of the Peninsula on price-to-rent, worse than the inner East Bay, and carrying a regulatory complication that none of the high-yield inland cities have to the same degree.
Related: Oakland, which has its own stabilisation ordinance and a higher gross yield; the high-yield East Bay for the other end of the region; and why published rental yields are wrong on the asking-rent problem in general.