At $2,409 in June 2026, the ZIP's Zillow ZORI gives the current typical observed asking-rent reading. It is a ZIP-level index blended across rental types, not a contract-rent series or a set of measured bedroom quotes. Annualizing that index and dividing it by the ZIP median sold price produces a 4.35% cross-source screening ratio. It is not a complete property-economics measure: it uses no unit-specific costs and does not arise from rental transactions. The central tension is therefore a current asking-rent reading alongside resale stock that has changed materially, not evidence that the rental and for-sale series describe one market.
The longer direct Zillow ZIP history supports a stable-growth description, but its pace has changed. Exact same-month annualized changes were 2.52% over 1 year, 1.80% over 3 years, and 4.15% over 5 years. Recent positive movement thus confirms the broad upward path while breaking from the faster five-year pace; it also accelerated relative to the softer three-year pace. The history has 100% coverage. The 2.62% annualized monthly-return variability sets a scale for the confidence appropriate to one current snapshot. Separately, maximum drawdown reached -3.06%, a historical retreat rather than a guarantee. Among history-eligible ZIPs, the transparent national discovery stability rank was 919, the balanced rank was 1,041, and the momentum rank was 1,417, where lower ranks are higher. These backward-looking measurements are neither forecasts nor investment recommendations.
Redfin's direct rolling-three-month ZIP resale observation is a separate for-sale record, not rental evidence. It reports a $664,850 median sold price, up 2.28% year over year, across 184 homes sold; median marketing time was 34 days. Inventory stood at 161 homes after a 32.19% annual increase, while months of supply measured 2.7. That measure sets available resale inventory against the recent sales pace, indicating relatively limited resale supply at that pace rather than rental availability. The average sale-to-list ratio was 100.72%, and 38.58% of sales closed above list. Those sale outcomes and the supply reading coexist with inventory growth. This challenges a smooth extension of the rent-history path into property economics, even though the resale price and ZORI each increased over the year.
The 21401 label serves both as Zillow's ZIP market identifier and as a match to a Census ZCTA. A ZCTA is a statistical area and is not identical to a USPS delivery ZIP, a boundary distinction that matters before applying area statistics to an address. The matched ACS 2024 five-year survey places median gross rent at $2,150; it covers occupied renter homes and includes selected utilities. That rent concept differs from ZORI rather than providing an interchangeable asking-rent measure. HUD's local two-bedroom FMR/SAFMR standard is $1,857, an administrative bedroom-specific standard rather than asking rent. Scaling ZIP ZORI using the local HUD ladder yields modelled monthly estimates of $1,767 for a studio, $1,960 for one bedroom, $2,409 for two, $3,059 for three, and $3,387 for four. They are modelled estimates, never measured bedroom rents.
Income and burden create another constraint on a headline rent. At the stated index, a 30% required-income screen equals $96,360 annually. This is arithmetic only—not advice and not an applicant-qualification rule. The ACS ZCTA median household income is $123,381, so the screen sits below that all-household benchmark, but median household income is not a renter-income distribution. In the same ACS survey, 43.7% of renter households were estimated to have rent burden at that threshold. The coexistence of that burden measure with the mechanical screen cautions against treating a ZIP-wide median income as proof of affordability for any household or unit. Because ACS gross rent includes selected utilities and is a five-year occupied-home survey, it cannot be substituted for a current listing's bill.
The ACS ZCTA housing picture is broader than either index. Of 19,567 housing units, 1,583 were vacant, an 8.1% vacancy rate; 299 of the vacant units were classified as for rent. Vacancy therefore is an area-level status count, not proof that a particular rental is available, competitively priced, or suited to a specified household. The stock count included 13,164 single-family units and 3,408 large multifamily units. Those categories help describe composition, but they do not identify bedroom mix, condition, utility terms, or the supply of current listings. The occupancy, stock, and vacancy measures remain survey estimates for the matched ZCTA rather than a census of individual properties.
For wider context only, the Annapolis city scope rent was $2,398, the Anne Arundel County scope rent was $2,370, and the Baltimore–Columbia–Towson, MD metro scope rent was $1,936. The direct ZIP ZORI is closest to the city and county context readings and materially separated from the metro context reading. These are wider-geography context values, not substitutes for the ZIP asking-rent index, the matched ZCTA survey, or the local HUD standard. The separation should be read only as scope-specific context because city, county, and metro reporting areas differ from both the direct ZIP and the statistical ZCTA. It does not validate a particular asking rent, describe a property's location within a delivery ZIP, or establish a tenant's budget.
Each evidence stream has a different date and construction: ZORI is observed asking-rent indexing, ACS is a survey of occupied renter homes, HUD is an administrative standard, and Redfin is direct ZIP resale activity. A property-level use would require confirming the advertised rent, bedroom count, included utilities, lease term, concessions, and live availability rather than relying on the modelled ladder or area vacancy. It would also require matching the address to the applicable boundary and checking a sale's listing status, closing date, property type, condition, and list-price record before pairing it with the resale figures. No unit-specific lease transactions, property expenses, financing terms, or operating records appear here. Can the relevant unit details be documented well enough to keep these aggregate screens in their proper roles?