With the floor area of 135 Camberwell resolved, the open question changed: does the $1.8M buyer in Robina pay for 197 m² split between house + granny flat the way they pay for 205 m² under one roof? Search on realestate.com.au — Robina 4226, 4+ bedrooms, keyword filter granny flat: 25 results since Jan 2025, ~18 in the last 12 months. Every relevant candidate was opened to separate a real granny flat from "granny flat potential", a distinction the filter does not make and that changes the whole reading.
| Sale | Price | Land | House | $/m² house | What it actually has |
|---|---|---|---|---|---|
| 13 Miling Ct · Jun 2025 | $1,845,000 | 800 m² | 371 m² | $4,973 | Self-contained guest room inside the house · golf frontage |
| 115 Newport Dr · Jan 2026 | $1,775,000 | 822 m² | 354 m² | $5,014 | REAL self-contained studio. 5 bedrooms + studio, 3 bathrooms. Land almost identical to ours — and 80% more house. It is the ceiling of the cohort. |
| 316 Ron Penhaligon Way · Jul 2025 | $1,735,000 | 737 m² | 337 m² | $5,148 | Granny flat only potential — "could easily be converted" |
| 54 Olympus Dr · Sep 2025 | $1,660,000 | 659 m² | 335 m² | $4,955 | Dual living with plumbing in place; second kitchen already removed |
The four trade between $4,955 and $5,148 per m² of house — a 3.9% band, CV of 1.5%, the tightest band in the whole Robina comp set. Applied to our 197.25 m² it would give $977k–$1,016k, which does not mean Springvale16 is worth $1.0M: it means this cohort is not our comp set. A 350 m² house carries a low $/m² because the building dominates; a 200 m² house carries a high $/m² because the land dominates — 135 Camberwell does $8,759/m². The granny flat does not move us into the $1.75M+ band, because what defines that band is the 350 m² house, not the granny flat.
| Sale | Price | Land | $/m² land | What it has |
|---|---|---|---|---|
| 19 Frankston Ct · 18 Nov 2025 | $1,570,000 | not published | — | 4 bedrooms, 2 bathrooms, renovated pool (new deck, frameless glass fence), new kitchen, cul-de-sac — and the granny flat is a CONVERTED GARAGE. The exact origin of ours. The P50 of our twin comps with no premium is $1,575,000: a $5,000 difference. |
| 7 Cremorne Ct · Oct 2025 | $1,550,000 | 881 m² | $1,759 | The ceiling of the income argument: 4 bedrooms + TWO self-contained units with kitchenette and private entry, 4 bathrooms, saltwater pool. The most dual living a normal Robina lot can carry — and 18% below the $/m² of land our break-even needs. |
135 Camberwell delivers 205.5 m² under one roof. Springvale16 delivers 143.37 m² of main house + 53.88 m² of separate granny flat. The total matches; the product does not. At the twin's rate ($8,759/m²), the main house is worth $1,255,796 — leaving $476,761 for the granny flat to carry to break-even, or $8,849/m² of granny flat, above the $/m² of the twin's own premium house. Crediting the granny flat at a defensible value ($200–280k), the total lands at $1.456M–$1.536M.
| Method | Result |
|---|---|
| P50 of twin comps, no location premium (n=5) | $1,575,000 |
| 135 Camberwell rate on the main house + granny flat at a defensible value | $1.456M – $1.536M |
| $/m² of land yardstick, no product premium | ~$1,575,000 |
| 19 Frankston Ct — the product mirror, converted-garage granny flat | $1,570,000 |
| Break-even sale price (EAC 21 Sep) · The Deal | $1,762,153 |
Answer: no. The $1,850,000 thesis sits 4.2% above the ceiling of the whole granny-flat cohort, with 44% less internal area. That does not prove the granny flat is worth nothing — dual living is a paid-for attribute on the Gold Coast and the granny flat earns rent (≈$450–550/week), which counts in a hold scenario and does not show up in a flip. It proves the granny flat is not the path to $1.85M on an 806 m² lot with 143 m² of main house. The path to $1.85M, if it exists, is 135 Camberwell's — continuous house, multiple living zones, premium finish — and not even that one cleared $1.8M.
For G5: (1) take 19 Frankston Ct as the central comp of the ARV discussion; (2) reconsider flip vs hold — the granny flat is worth more as income than at resale, and if the vehicle becomes an income one the Reno OS has to be reassessed on yield, not on margin over GDV; (3) do not spend reno money trying to make the granny flat look more like house; (4) ask for the land size of 19 Frankston, not published.
The $/m² of house lens is unstable (our floor area has already gone from 224 to 197.25 m²) and does not exist for decisive comps such as 28 Merion and 173 Ron Penhaligon. The land lens exists for all of them. Our lot: 806 m².
| Reference | Value on our lot | $/m² of land | vs neighbour |
|---|---|---|---|
| 17 Springvale St — next door, sold 17 Sep | $1,134,046 | $1,407 | base |
| 7 Cremorne Ct — the most dual living | $1,417,554 | $1,759 | +25% |
| P50 of twin comps, no location premium | $1,575,000 | $1,954 | +39% |
| Break-even sale price (EAC 21 Sep) · The Deal | $1,762,153 | $2,186 | +55% |
| 135 Camberwell applied to our lot | $1,815,918 | $2,253 | +60% |
| Team thesis | $1,850,000 | $2,295 | +63% |
The number of the week: the break-even needs 92% of all the product premium the market has already shown it pays in this band. The thesis needs 105% — more than Robina's best finished house achieved. And 17 Springvale does not prove we paid too much: it is 3 bedrooms, no pool, with an openly motivated seller. It proves the floor of our street is $1,407/m².
Crawl of 15 Sep over Robina 4226, 4+ bedroom houses: 46 sales with a price between Mar and Sep 2026 and 6 active listings above $1.4M in the relevant format. The 3 Sep baseline had left out 10 sales in the $1.43–1.91M band; they are in now. Since 3 Sep, a single new sale has been recorded — 4 Beauty Point Dr, closed on 2 Sep, price not disclosed.
| Indicator | 3 Sep (W36) | 15 Sep (W38) | Reading |
|---|---|---|---|
| Internal area of the asset | 224 m² (to be measured) | 197.25 m² | FRD plan: 143.37 house + 53.88 granny flat. Cuts the $/m² base by 12% |
| Buy-box median (4+ bed, 600–1,100 m², 6 months) | $1,800,000 | $1,700,000 | Baseline skewed upward: the $1.43–1.67M sales were missing |
| Buy-box median · last 90 days | — | $1,650,000 | −5.3% vs the previous 90 days ($1,742,500) — part mix, part registration lag |
| P50 of twin comps, NO location premium (n=5) | $1,600,000 | $1,575,000 | Honest base if the Form 2 does not deliver the reserve |
| P50 of twin comps WITH a premium (water/golf/reserve, n=3) | — | $1,700,000 | Konda $1.715M · Leopardwood $1.700M · Emerton $1.655M |
| Sales ≥$1.8M with house ≤250 m² | 2 | 1 + 2 to confirm | Thesis validation signal still not triggered |
| Listings asking ≥$1.7M in the 4/2/2 format | 0 | 4 | Pine Valley $1.75M · Pinnacle $1.759M · Protea $1.75M (reduced) · Auriga $1.949M |
| 31 Lancewood Cct (renovated twin, 260 m²) | $1.6M+ | $1.6M+ | 12 days since first capture, no reduction and no sale |
Next price-discovery event: six auctions in Robina between 18 and 26 Sep (44 Kirralee, 27 Ron Penhaligon, 26 Olympus, 5 Unley, 30 Manly, 146 Camberwell — the last three in 4/2/2). Two weeks of spring campaign are worth more than any estimate: the next snapshot (22 Sep) must record those results before any scope-upgrade decision.
All-in cost of the "Presented" spec ($270k): purchase + duty $49,698 + legals $8k + reno + 9 months of holding at $7,007/month + 2.75% commission + $7.5k marketing = $1,646,541. Above the P50 of the twin comps. Marcos's thesis — ARV $1.85M for the quality of the land and access to a restricted public reserve maintained by council — sits 2.8% above the observed ceiling of the twins ($1.485M–$1.80M) and 15.6% above P50. It requires the comp set not to contain the differentiator.
| Spec | Reno | Break-even | Profit @ P50 $1.60M | Profit @ thesis $1.85M |
|---|---|---|---|---|
| Presented — the cheapest earns the most | $270,000 | $1,646,541 | −$45,260 | +$197,864 · IRR 44.4% |
| A+ surgical | $315,000 | $1,692,813 | −$90,260 | +$152,864 |
| Premium | $375,000 | $1,754,510 | −$150,260 | +$92,864 |
| Premium+ | $420,000 | $1,800,783 | −$195,260 | +$47,864 |
Why the cheap spec is the optimum: if the premium comes from the land and not the finish, every extra dollar of works is a subtraction. Asymmetry: thesis right + cheap spec = +$198k; thesis wrong by $150k + expensive spec = −$4k. Robust decision: keep the budget lean and test the thesis. Three tests, all before G5 (12 Sep):
Undefined between 193, 224 and 240 m². If it comes out at 193 m², the twins' $/m² points to an ARV of $1.35–1.55M — below break-even on any spec — and the conversation becomes hold-and-rent. Gate ahead of any budget.
Binary. Easement registered on title: transfers, can be advertised, real premium. Informal council tolerance: does not transfer, worth zero in the buyer's bank valuation, and advertising it is a risk under the ACL. Also check the environmental/bushfire overlay of the adjoining land — it can cut ARV instead of adding.
Every sale ≥$1.7M in the area has a double garage; without it the premium scenario does not close. In R1 (14–25 Sep), ask without leading: "what does this house do because of the reserve?". Three agents saying $1.8M+ = licence to spend; nobody mentioning it = $105k saved. And count how many twins in the crawl share the same adjacency — if 2–3 do, the premium is already in the comp set.
Internal area measured on the FRD plan: 197.25 m² (143.37 house + 53.88 granny flat), not the model's 224 m². $1,841,000 ÷ 197.25 m² = $9,333/m² — 62% above the median of Robina sales ($5,755/m²) and within 4% of the highest $/m² in the sample. A single sale in six months closed above that, and it was a 170 m² house on a 987 m² block: land value. It is exactly the land thesis's argument: if Springvale is worth $1.85M, it is for the lot and the adjacency, not the built m².
Each dot is a Mar–Sep 2026 sale with a known floor area. The shaded zone is where the model needs Springvale to sell — and where the market has put no house of this size.
✓ present · – absent · ? to confirm with Rei/Emma. Ceiling height, orientation and privacy cannot be bought in the reno: they are the three questions that decide whether $1.8M is reachable. The "Privacy / reserve / golf" column is where the 5 Sep thesis lives — 28 Merion and 28 Leopardwood sold that.
Twenty-nine sales above $1.7M in six months (≈4.8/month). Every one that cleared $1.87M has a house ≥330 m², water or golf frontage, dual living, or a signature reno — sauna, bar, fireplace, microcement, putting green.
The only two 4/2/2s of 220–250 m² that crossed $1.8M sold different things: 9 Rosebud sold full reno + high ceilings + north-facing deck; 28 Merion sold location (end of court, gated, next to the golf course).
6 Yowani (exactly 224 m², pool, presented) closed at $1.575M. 341 Ron Penhaligon (246 m², renovated kitchen and two bathrooms, Robina Woods) at $1.485M. 28 Leopardwood (229 m², Rosebud-grade reno + reserve at the rear) at $1.70M. And 31 Lancewood — 260 m², complete, with a custom bar — is asking $1.6M+.
To target >$1.85M with confidence the house would need to become ≥300 m² or dual living: that is small development (SPV, DA, time, capital), not Fix & Flip. The alternative is the land thesis — which only holds if the Form 2 and the agents confirm it.
| Address | Status | Price | $/m² house | Bed | Bath | Car | Land | House | Date | Why it is worth that |
|---|
| Isolated attribute | Comp pair | Δ price | Reading |
|---|---|---|---|
| Full reno vs "presented" | 9 Rosebud (221 m², full reno, high ceilings, N deck) vs 6 Yowani (224 m², pool, presented) | +$225k | The maximum premium the reno bought in Robina for this size. Includes ceiling height and a north-facing deck — two items Springvale may not have. |
| Partial reno (kitchen + 2 bathrooms) | 341 Ron Penhaligon (246 m², granite kitchen, new bathrooms) vs 6 Yowani (224 m², no reno) | −$90k | A "halfway" reno does not pay: the market did not tell it apart from a merely well-presented house. Either do it all, or don't. |
| Backing onto reserve (privacy) | 28 Leopardwood (229 m², Rosebud-grade reno, reserve) vs 9 Rosebud (221 m², reno, no reserve, 865 m²) | −$100k | Smaller land (648 vs 865 m²) cancelled the reserve. Land matters more than a rear outlook in this band. |
| Signature location (golf/court/gate) | 28 Merion (840 m², no reno declared, golf) vs 6 Yowani (760 m², presented) | +$235k | Location paid the same as a full reno — without spending $400k. Not available to Springvale. |
| +100 m² of house (330 vs 229) | 22 Lacerta (330 m², park, new kitchen, no pool) vs 28 Leopardwood (229 m², reno, reserve, pool) | +$475k | Size paid ~$4.7k per additional m². Extending 80–100 m² at $4.5–6k/m² breaks even or gains little — only makes sense if it unlocks the $2M+ band. |
| Size without quality | 16 Collingwood (436 m², 1,074 m², pool) vs 1 Julatten (366 m², 912 m², stone, timber) | −$500k | Raw floor area without finish sells at $3.9k/m². Size only pays with a matching level of finish. |
| Signature reno | 173 Ron Penhaligon (799 m², sauna, bar, fireplace, microcement, putting green) vs 9 Rosebud (865 m², "clean" full reno) | +$100k | The "hero feature" is worth ~$100k above a full reno — but costs part of that. Marginal, not structural. |
Costs are Gold Coast 2026 estimates for validation with Rei — they serve to rank priorities, not as a budget.
| Attribute | In the ≥$1.8M sales | Can it be bought? | Est. cost | Effect on ARV | Priority |
|---|
Full reno within the existing plan: stone island kitchen + induction, 2 full-height bathrooms, multi-zone ducted AC, continuous flooring, north-facing deck with bifolds and blinds, pool + compliance, solar, façade and landscaping. No extension.
Reno A + an 80–100 m² extension (second living room / master wing / media room). Changes the house's category and unlocks the band where there were 9 sales in 6 months.
Lean Reno A + a secondary dwelling on the 806 m² lot (self-contained 60 m²). Attracts investors and multigenerational families; wide range because it depends heavily on the finish.
Break-even = purchase $1.203M + entry/holding/sale ~$210k + reno. None of the three paths has a P50 above break-even on today's comps; A is the only one where three non-purchasable attributes (ceiling height, orientation, privacy) can turn the game without extra capital.
1. Ceiling height. Rosebud, Leopardwood and 173 RP have "soaring / high / raked ceilings". If Springvale has a standard 2.4 m, path A loses ~$100k of ceiling and there is no way to buy that without touching the roof.
2. Backyard orientation. Rosebud sold a "north facing deck"; Leopardwood "northeast entertaining". A south-facing deck in Robina is a discount, not a premium.
3. Pool privacy. Merion sold "no neighbouring homes overlooking the pool". A two-storey neighbour looking onto the pool takes Springvale off the list of who pays $1.8M.
Replace ARV $1.841M with a base case of $1.65M (P50) and an upside of $1.80M (P90) conditional on the three answers above. Cost path A in detail with Rei before any demolition; discard B and C as Fix & Flip — if either makes sense, it is a new project, in its own SPV, with a development thesis and a margin ≥25% over GDV.
Meanwhile, monitor weekly: new 4/2/2 listings on 700–900 m² in Robina, days on market and price reductions — the $1.8M thesis only gains strength if a third sale of a ≤250 m² house appears above it.