The Property Brief — what the evidence shows. You draw the conclusion.
Confidence: Low — based on 21 comparable auction sales (0 of this exact configuration) · model runs ±33% for this configuration · confidence held at Low — under 20 same-configuration sales · thin comparable support for this configuration — read the range as wider, not as a warning · the model's typical miss is ~11% against published results — every hit and miss
Suburb risk flags
crime index: 67 — above sydney average · trains ~every 10 min at peak
Where every number comes from
- Fair value: AI PropertyEye price model 2026-08-11-v5 — trained on 146,492 published auction sales; configuration-level (suburb × type × beds), not an appraisal of this exact home
- Risk flags: suburb-level overlays (BOCSAR crime, flood/bushfire intel, aircraft bands, GTFS transit) — NOT a per-parcel assessment
Rent evidence — asking vs transacted, Miranda · postcode 2228
$680/wk median live asking rent for 1-bed homes in Miranda (3 live rental listings on agency websites, this week's sweep).
$595–$680/wk is what 1-bed units in 2228 actually transact for (middle half of 31 bonds lodged, Q2 2026 · median $640/wk — NSW Fair Trading, CC BY).
2019Q1 → 2026Q2: $500 → $710/wk · +5% over the last 12 months (latest four quarters averaged against the previous four) — quarterly new-tenancy medians for units in 2228, all bedroom counts pooled.
Asking rents are what agents hope for; bond figures are what tenants actually signed at (NSW Fair Trading bond lodgements, CC BY). No yield % is derived here: each rent above is a different grain from this home's price, and a ratio of two grains reads as precision it doesn't have. New-tenancy rents run hotter than sitting rents. Not financial advice.
Growth trend — Miranda units at auction
Not enough Miranda unit auctions on record to show a trend — this fills in as the archive grows.
Suburb-level, from published auction results. No single building has enough sales for a block-level series, so this is the honest grain; sales-mix shifts move raw medians, and the measured rate (when shown) controls for bed count.
Published auction results at this address
| Auction week | Result | Price | Agency |
|---|---|---|---|
| 8 Aug 2026 | Withdrawn | withheld | Highland Sutherland Shire and St George |
From Domain's published weekly auction results — individual outcomes as published; withheld prices stay withheld.
Agencies' own published sales nearby — Miranda, around 1 beds
| Address | Sold | Price | Beds | Agency |
|---|---|---|---|---|
| 49/7-15 Jackson Avenue | by 26 Aug 2026 | $840,000 | 2 | Laing+Simmons |
| Bb601/18 University Road | by 16 Aug 2026 | $710,000 | 1 | Laing+Simmons |
| 50/8-12 Wandella Road | by 16 Aug 2026 | $860,000 | 2 | Laing+Simmons |
| 2/25-27 Kiora Road | by 16 Aug 2026 | $896,500 | 2 | Laing+Simmons |
Prices as published on the selling agency's own website (their sold pages) — includes private-treaty sales the auction results never see. Only sales the agency chose to publish a figure for appear here.
↩ Withdrawn from auction — what happens next
Withdrawn from auction isn't the end of a sale — of withdrawn Sydney homes that went on to sell, half sold within 65 days.
63% sold within three months of the scheduled auction.
1,535 withdrawn-then-sold homes across Sydney. We don't publish a price figure for withdrawn homes: the measured gap is too unstable to stand behind, and a number we can't defend is worse than none. Based on withdrawn homes that later sold. Homes that never sold aren't in the figure — withdrawal sometimes means a private deal was already done, and sometimes means the listing simply ended.
Sale history — the official settled-sale record at this address
| Exchanged | Price | Change since prior sale |
|---|---|---|
| 14 Apr 2025 | $625,000 | — |
Settled-sale records from the NSW Valuer General's Property Sales Information. Dates are contract (exchange) dates; sales appear only once lodged, so the most recent months are incomplete and revise upward — never read the latest month as final. Two entries on one date are usually a single transaction recorded across parts. © State of New South Wales (Valuation NSW).
Construction risk — what no listing tells you
Building Commission NSW / Strata Community Association 2023 Strata Defects Survey: 53% of newly registered strata buildings reported at least one serious defect in common property (39% in the 2021 survey). Defects are common, not exceptional — and new does not mean safe (Opal Tower was under a year old).
Get an independent structural and strata inspection before you commit — the defect base rate makes it the single highest-value dollar in the purchase.
We cannot see inside a building — no one can from public records. This section states base rates and observable adjacency only, and will never call a building safe.
The building — its registered strata scheme
Strata plan SP95544 registered 2019 — the building is at least 7 years old · 62 lots in the scheme.
Registration is an age floor, not a build date — a converted building is older than its strata plan. Scheme records from the NSW Strata Hub. © State of New South Wales (Spatial Services, a business unit of the Department of Customer Service NSW). For current information go to spatial.nsw.gov.au.
School catchments
Miranda PS (primary) · Port Hacking HS (secondary)
6 primary schools within 2 km · ICSEA ~1067 — the socio-educational advantage of the school community (My School, national average ~1000): a demographic measure, never a results or quality score.
Suburb-centre approximation for Miranda — addresses near a boundary can differ. Verify the exact address at School Finder. Intake zones cover government schools only.
Want the numbers agents don't volunteer?
Check any Sydney suburb free — Market Mint market-conditions readout, live median, clearance trend and the price model — at aipropertyeye.com. No sign-up for the suburb check; the full board is 7 days free.
Data from published auction results (Domain) and live portal listings. Estimates are never valuations or financial advice. Own this home and want a result corrected? hello@aipropertyeye.com.