The immediate tension in 94117 is a very high current asking-rent reading after unusually fast recent movement, against a longer series that also rose but more slowly. Zillow’s June 2026 ZIP ZORI is $4,334 per month, a typical observed asking-rent index blended across rental types rather than a quote for any one dwelling. The exact same-month gain was 22.6% over one year, while annualized same-month changes were 9.7% over three years and 7.6% over five years. Those are backward-looking index measurements, not forecasts or investment recommendations. The 94117 label is both a Zillow ZIP market identifier and a Census ZCTA match; a ZCTA is a statistical area and is not identical to a USPS delivery ZIP.
The historical record supplies unusually complete evidence but not a guarantee that one current reading is representative of every listing. Coverage is 100%, with 121 monthly observations and 120 consecutive monthly returns. Monthly rent changes annualized to 3.5% variability, so the path was not perfectly smooth despite the strong recent gain. Separately, the maximum drawdown was 13.2%, identifying the largest historical peak-to-trough decline in the index. Transparent national discovery ranks among history-eligible ZIPs were 12 for momentum, 2,255 for stability, and 527 for the balanced measure, where lower ranks are stronger. The recent acceleration confirms the longer upward direction, but its pace breaks from the more moderate multi-year path. That instability warrants less confidence in treating a single current ZORI snapshot as a durable rent level.
Cross-source comparisons are useful only when their different universes remain visible. The matched 2024 ACS five-year ZCTA reports median gross rent of $2,881 for occupied renter homes; gross rent includes selected utilities and is not an asking-rent measure. ZIP ZORI is therefore 50.4% above that survey median, a difference that can reflect tenant occupancy, timing, utilities, and measure design rather than a direct rent-comparable gap. The FY2026 local HUD two-bedroom FMR is $3,604, placing ZORI 20.3% higher. HUD FMR is an administrative, bedroom-specific standard, not observed asking rent, and neither the ACS estimate nor FMR should be substituted for a current advertised unit price.
The bedroom figures are modelled monthly ZIP estimates, not measured bedroom rents. They scale the ZIP ZORI by the local HUD bedroom ladder: $2,988 for a studio, $3,580 for one bedroom, the current index level for two bedrooms, $5,537 for three bedrooms, and $5,739 for four bedrooms. This is a sizing framework, not evidence that available units rent at those exact amounts. Applying a 30% of gross-income screen to the ZIP ZORI produces required annual income of $173,360. That is arithmetic, not advice and not an applicant qualification rule. It exceeds the ACS median household income of $167,066, while the ZIP asking-rent-to-income comparison is 31.1%. Household income is also broader than renter income, adding another reason not to read this screen as a household-specific affordability determination.
The ACS ZCTA describes a renter-heavy stock with meaningful vacancy, but neither measure proves availability or cost for a particular unit. Of 19,801 housing units, the reported vacancy rate is 11.3%; renter households account for 71.2% of occupied homes. Among renter households, 32.4% report spending at least 30% of income on gross rent. At wider scope only, the City of San Francisco context has a $4,401 rent index, San Francisco County context matches that City figure, and the San Francisco-Oakland-Berkeley, CA metro context has a $3,301 rent index, 3.3% apartment vacancy, and 19.7 apartment days on market. These city, county, and metro values are comparison context, not substitutes for ZIP-level conditions or evidence about an individual building.
The direct rolling-three-month ZIP resale observation ending June 2026 shows a for-sale market with rapid turnover indicators, not rental transactions. Median sold price was $1,875,076, up 10.3% year over year, with 90 homes sold and a median 17 days on market. The inventory reading was 14 homes and months of supply stood at 0.5. Sale-to-list signals were also elevated: the average sale-to-list ratio was 120.1%, and 83.0% of sales closed above list price. Annualized ZIP ZORI divided by median sold price produces a 2.8% cross-source screening ratio only; it is not a measure of property economics. The resale evidence moves in the same general direction as the rent-history acceleration, yet the income screen remains a counterweight because the ZIP-level asking-rent benchmark sits above the median-household-income-based 30% threshold.
Several limits prevent these data from becoming unit-level conclusions. ZORI blends rental types and does not disclose the condition, lease term, furnishings, concessions, utility package, or bedroom mix of a specific listing. ACS is a five-year survey of occupied renter homes with sampling uncertainty, while HUD is an administrative standard with a different purpose. The ZCTA-to-ZIP match supports this geographic comparison but does not turn a statistical boundary into a USPS delivery geography. Redfin’s observation is direct ZIP resale evidence, but it supplies no rental comparable, operating expense, financing, tax, maintenance, or tenant information. None of the rent history, burden, vacancy, or resale measures establishes what a particular property will command or experience.
Property-level checks needed to apply this ZIP report are concrete: the advertised asking rent, exact bedroom and bathroom configuration, square footage, included utilities, lease duration, concessions, furnishing status, available date, and whether the unit’s address actually falls within the relevant delivery geography. A resale comparison also needs the specific property’s sale record, list-price history, physical condition, and transaction timing rather than the ZIP median alone. For renter-occupied or vacant structures, the classification and availability should be verified directly rather than inferred from ZCTA totals. The central question is whether the identified unit’s documented terms resemble the broad asking-rent index and modelled bedroom ladder, not whether the ZIP average applies universally.