At $1,019,770, the 98117 median sold price sits beside a current $2,402 Zillow asking-rent index, creating a sharp rent-versus-resale tension rather than a unified property-economic measure. The direct ZIP Redfin resale price changed by 0.97% from a year earlier, while the annualized ZIP ZORI divided by that sold price produces a 2.83% cross-source screening ratio. That ratio is neither a cap rate, net return, expected return, nor property yield. Redfin is a direct rolling-three-month ZIP resale observation, whereas Zillow records rental asking-rent conditions; the figures can be screened together but do not describe the same transactions.
Rent history shows positive but decelerating movement. The one-year exact same-month annualized change was 1.24%, compared with 2.52% over three years and 4.34% over five years. Thus, the current positive direction breaks from the stronger longer-run growth pace rather than confirming it. Annualized monthly-return variability was 2.85%, which supports more confidence in the current ZORI snapshot than a highly erratic series would, though it does not eliminate listing-level variation. The worst peak-to-trough drawdown was -4.35%, a backward-looking reminder that the index has declined before. History has 100% coverage, and its transparent national discovery ranks are 1,597 for momentum, 1,334 for stability, and 1,569 for the balanced measure, with lower ranks indicating higher placement among history-eligible ZIPs. These are descriptive discovery measures, not forecasts or investment recommendations.
Current-rent comparisons require strict source separation. Zillow ZORI for 98117 is a typical observed asking-rent index blended across rental types, not a survey median or an administrative standard. The five-digit label 98117 is both Zillow's ZIP market identifier and its Census ZCTA match; a ZCTA is a statistical area and is not identical to a USPS delivery ZIP. In the Seattle city context, Zillow rent is $2,224; in the King County context it is $2,330; and in the Seattle-Tacoma-Bellevue, WA metro context it is $2,269, so the ZIP index is above each broader benchmark. The ACS five-year ZCTA survey reports $2,178 median gross rent for occupied renter homes, including selected utilities, or 10.3% below ZORI. The local HUD two-bedroom FMR/SAFMR standard is $2,501, placing ZORI 4.0% below that administrative bedroom-specific standard, which is not asking rent.
The bedroom ladder is useful only as a modelled translation of the all-rental-type ZIP index. Scaling the $2,402 ZORI by the local HUD ladder yields modelled monthly estimates of $1,992 for a studio, $2,061 for one bedroom, $2,402 for two bedrooms, $3,142 for three bedrooms, and $3,695 for four bedrooms. These are modelled estimates, never measured bedroom rents or observed unit-level asking quotes. The local HUD ladder supplies the proportions behind the calculation, but HUD FMR/SAFMR remains an administrative standard rather than evidence that available units ask at those amounts. A specific listing can depart from the ladder because ZORI is blended across rental types and the model does not observe unit condition, lease structure, utilities, or concessions.
The income and burden measures present a separate affordability screen. The ZCTA's ACS median household income is $182,984. Applying the 30% arithmetic screen to current ZORI gives required annual income of $96,080, and the current asking-rent-to-median-income calculation equals 15.8%. This is arithmetic, not advice and not an applicant qualification rule. ACS estimates that 1,150 of 4,103 renter-occupied households, or 28.0%, paid at least 30% of income toward gross rent. That burden statistic concerns surveyed occupied renter homes and cannot prove the cost burden of a particular available unit. For wider context only, the Seattle city-context burden share is 43.9% and the King County context share is 46.8%; those wider geographies should not be substituted for the ZIP ZCTA estimate.
Housing-stock evidence indicates a comparatively owner-heavy matched ZCTA, but it does not identify the availability or terms of a specific rental. ACS counts 15,725 housing units, of which 14,983 are occupied and 742 are vacant, for a 4.7% vacancy rate. Renter-occupied homes represent 27.4% of occupied units, and 388 vacant units are classified as for rent. Those counts help frame the aggregate stock and vacancy picture behind the rent index, but vacancy is not proof that a given unit is obtainable, appropriately priced, or comparable to the ZORI basket. Nor does the renter share establish demand, lease-up conditions, or a likely rent outcome for any individual property.
Redfin's direct ZIP resale signals depict brisk for-sale liquidity, not rental transactions. Its rolling-three-month observation records 178 homes sold, a median 6 days on market, and inventory of 83 homes. Months of supply stood at 1.4. The average sale-to-list result was 102.96%, while 45.71% of sales closed above list price. These resale indicators sit uneasily beside the slower one-year rent-index growth: quick marketing and above-list sales challenge any attempt to infer a weak for-sale market from rent deceleration, while the modest rent change prevents the resale evidence from being read as confirmation of accelerating asking rents. The tension is descriptive because Redfin sales and Zillow asking rents occupy different market universes.
The remaining uncertainty is property-specific. ZORI is an index rather than a listing comp, ACS is a five-year survey rather than a current offer sheet, HUD is a standard rather than an asking-rent survey, and Redfin describes completed resale activity rather than rental economics. Concrete checks should therefore verify the exact geography, bedroom and bathroom count, lease term, move-in timing, furnishing status, concessions, included utilities, active availability, and the date of the advertised rent. For a resale comparison, verify the relevant property's actual list price, sale date, condition, and nearby closed-sale selection rather than applying the ZIP median mechanically. The unresolved decision question is whether an individual unit's documented terms resemble the index and model inputs at all.