Rent and resale evidence point to a clear measurement tension in ZIP 90004. In June 2026, Zillow ZORI, a typical observed asking-rent index blended across rental types, was $2,331 per month and 2.34% lower than a year earlier. Yet the direct Redfin ZIP resale observation reported a $1,747,605 median sold price, 15.74% above its prior-year reading. Annualized ZIP ZORI divided by that sale price is 1.60%, a cross-source screening ratio only; it is not a cap rate, net return, expected return, or property yield. The five-digit 90004 label is both a Zillow ZIP market identifier and a Census ZCTA match. A ZCTA is a statistical area, not an area identical to a USPS delivery ZIP.
Against wider rent context, the Los Angeles city context value was $2,773.19, the Los Angeles County context value was $2,808, and the Los Angeles-Long Beach-Anaheim, CA metro context value was $2,927. Each is a city, county, or metro context rather than a ZIP-level substitute, and none converts into a unit-level comparable. The ZIP index therefore sits below all three broader rent benchmarks, while still describing asking rents rather than signed leases, occupied homes, or administrative payment standards. This separation matters because the lower ZIP reading cannot establish that a given available apartment is cheaper, nor can it explain the gap between rental and resale signals.
The backward-looking Zillow rent path supports the cooling label but not a claim about what comes next. Exact same-month annualized change was -2.34% over one year and -0.54% over three years, after a +2.08% annualized five-year change. Recent direction thus breaks from the longer positive path while extending the more recent softening. The monthly-return series has 2.50% annualized variability, which reduces the confidence that should be placed in a single current rent snapshot. Separately, maximum drawdown reached 4.93%, showing a historical retreat rather than a stable straight line. Coverage was 100% across 122 observations and 121 consecutive monthly returns. Transparent national discovery ranks among history-eligible ZIPs were 2,769 for momentum, 692 for stability, and 2,251 for the balanced measure, with lower ranks higher. These are measurements, not forecasts or investment recommendations.
The ACS 2024 five-year survey for the matched Census ZCTA reports median gross rent of $1,831. That measure covers occupied renter homes and selected utilities, unlike an observed asking-rent index, and it is $500, or 27.3%, below the current Zillow figure. The difference is a source-universe difference before it is evidence of any particular property’s pricing. HUD FY2026 sets a two-bedroom fair-market-rent standard of $3,070 for the local ladder. HUD FMR or SAFMR is an administrative, bedroom-specific standard rather than asking rent; it should not be treated as a rent quote or as an ACS replacement. The three series answer different questions on different populations and timing.
Using the proportions in that local HUD ladder to scale the ZIP ZORI produces modelled monthly estimates, not measured bedroom rents: $1,808 for a studio, $1,927 for one bedroom, $2,331 for two bedrooms, $3,066 for three bedrooms, and $3,548 for four bedrooms. Their pattern is useful for comparing bedroom sizes only within the modelled framework; it does not demonstrate actual available asking rents by bedroom. The 30% required-income screen calculates $93,240 in annual income at the ZIP asking-rent level, versus a $64,826 ACS median household income. The published asking-rent-to-income comparison is 43.1%. This is arithmetic, not advice and not an applicant qualification rule, and it does not measure any household’s actual rent payment.
Housing composition and survey burden make the aggregate renter backdrop important, but they do not identify a unit. Renter-occupied homes represent 83.85% of occupied housing in the ZCTA. The reported stock includes 9,068 units in large multifamily structures and 4,727 single-family units. Overall vacancy was 7.98%, with 992 units classified vacant for rent; neither figure proves availability, condition, concessions, or rent at a specific address. In the ACS renter sample, 58.8% of renter households reported paying 30% or more of income toward gross rent. That burden statistic is population-level survey evidence, not proof that a particular tenant or apartment faces the same cost pressure.
Redfin’s direct rolling-three-month ZIP resale observation supplies a separate liquidity and pricing readout. It recorded 35 homes sold, a 56-day median marketing time, and inventory of 90 homes, equivalent to 7.8 months of supply. The average sale-to-list ratio was 97.45%; 20.61% of sales closed above list. These are for-sale outcomes, not rental transactions, rental comparables, or property operating results. Together with the previously noted resale-price increase, the resale record challenges a simple reading that rent cooling is mirrored in all local housing measures. It does not resolve the tension: transaction volume, listing exposure, and sale negotiation say nothing directly about the rent of an individual home or apartment.
Several limits remain material. Zillow tracks a blended typical asking-rent index; ACS tracks surveyed occupied renter homes with selected utilities; HUD supplies administrative standards; and Redfin captures rolling ZIP resale activity. The ZCTA boundary and USPS delivery ZIP are not identical, dates differ across series, and the resale screen is not property economics. Property-level checks needed to interpret this packet are the actual bedroom count, current advertised rent and concessions, utility responsibility, lease terms, condition, listing availability, and the property’s own sale price, list price, marketing history, and transaction timing. The evidence neither forecasts rent or price nor supports a recommendation. The unresolved decision question is whether those property-specific facts align with the ZIP-level asking-rent cooling, survey burden, and distinct resale record.