No legal building on a Malé plot earns back its asking price
Malé plots are quoted at MVR 7,200–15,000 per square foot, yet an honest residual valuation — completed value minus every cost of building — supports a fraction of that, and is negative in our base case. The gap measures a legal regime in which holding is free, selling is hard, and the highest-value use of prime land is waiting.
The scarcest land, by law as much as by geography
Malé island’s four wards — Henveiru, Galolhu, Maafannu, Machchangolhi — held 137,238 people at Census 2022 on roughly 2 km², a density around 62,000–70,000 persons per km² [OFFICIAL census; OWN-CALC density, area basis varies by source]. The island added only about 9% population in the eight years to 2022 while the administrative city grew 37.7% — the growth all went to reclaimed land across the bridge [OFFICIAL].
| Ward | 2014 | 2022 |
|---|---|---|
| Maafannu | 42,788 | 47,036 |
| Henveiru | 31,391 | 34,012 |
| Galolhu | 25,788 | 28,484 |
| Machchangolhi | 26,002 | 27,706 |
| Malé island total | 125,969 | 137,238 |
Populations are Census 2022 final ward tables [OFFICIAL]. The economic point of the table is what does not exist beneath it: there is no ward-level price evidence anywhere in the public domain — no official, broker or media source publishes land values for Henveiru versus Galolhu versus Maafannu versus Machchangolhi. Everything that follows about “Malé land prices” is island-level evidence, and thin even at that.
The scarcity is now policy-locked, not merely geographic. The December 2025 Housing for All scheme — 15,000 plots of 1,250 sqft (116 sqm) and 7,900 flats — puts zero new residential land on Malé island; all supply is in Hulhumalé, Gulhifalhu, Giraavaru and RasMalé [MEDIA reporting official policy]. The scheme’s eligibility line is itself a price signal: owning no more than 400 sqft (37 sqm) anywhere still counts as landless [MEDIA] — an official admission that the inherited fragments generations of subdivision have produced are not viable homes.
What the evidence actually says a plot costs
No transaction register exists in the Maldives, so every land “price” below is asking, administered or listing tier — and the tiers disagree with each other in ways we carry rather than resolve.
Four markers matter. First, the state’s own administered sale of 19 Malé plots at a flat MVR 15,000/sqft (about MVR 161,000/sqm; roughly MVR 792m ≈ USD 51m total at the peg) — plots on Orchid Magu, Boduthakurufaanu Magu and west of Villa College [MEDIA, The Edition; vintage assessed 2018–19, article date unverifiable because the site was unreachable at research time]. A flat price across radically different micro-locations tells you the government itself does not price frontage or corners. Second, the structured-portal listing band of MVR 7,200–9,400/sqft for private Malé residential plots (n=3, July 2026) [ASKING]. Third, the only “sold” marker in the entire evidence base: a 425 sqft Maafannu fragment marked sold at MVR 4.0m — MVR 9,412/sqft (USD 259,000 total at peg) [LISTING, platform flag unverified]. Fourth, a broker commentary quoting prime Malé at MVR 12,000–20,000/sqft [INDUSTRY-EST]: our Greater Malé paper found that source internally inconsistent and 2.5–6× portal evidence and discarded it as probable unit confusion, while the land dossier records it as the quoted ceiling. We carry it here as a contested ceiling, not evidence — the conflict stands.
Two honest readings of this ladder are possible. If the administered 15,000 was a real 2018–19 clearing level, then 2026 portal asks of 7,200–9,400 imply nominal stagnation or decline — or the administered price was always above market and has anchored quoting behaviour since. Without transactions, these cannot be distinguished [OWN-CALC inference on conflicting tiers]. Either way the national gradient is extreme: outer-atoll beachfront lists at MVR 675/sqft — Malé quotes run 11–22× that, and a Hulhumalé plot asking MVR 10,651/sqft shows reclaimed land now quoting inside the old-island band [ASKING].
On frontage and corner premiums the record is empty: no quantified evidence exists anywhere public [GAP]. Market participants cite main-road frontage (ground-floor retail rents) and corner plots (light, ventilation, lettable efficiency) as real drivers [COMMENTARY]. In the residual model below we carry a frontage/corner sensitivity of +10–25% on land value as a labeled assumption [OWN-CALC], because no source can support a number.
The envelope: height rules, cores and the 600 sqft floor
What a plot is worth depends on what you may build on it, and here the capital’s rulebook is partly invisible. Malé island building permits run under the Construction Act (4/2017) and national building code; height is discretionary — “subject to plot location, area and usage” — and the numeric Malé height/FAR schedule exists only in a Dhivehi regulation we could not retrieve in English [OFFICIAL framework; GAP on the schedule]. Observed practice is 6–10 storey, party-wall-to-party-wall, near-100% coverage redevelopment [COMMENTARY]. The clearest published benchmark is next door: Hulhumalé’s development-control schedule, amended March 2026.
Under those tiers a plot under 400 sqft supports 1 storey; 801–900 sqft supports 8 storeys; over 901 sqft supports 10 storeys (30.5 m), with the ground floor surrendered to parking and the first floor to non-residential use above 18.5 m [OFFICIAL via MEDIA]. Whether Malé island administratively mirrors this is unconfirmed [GAP] — but the step-function logic visibly governs Malé outcomes anyway, because it compounds with lift-core arithmetic: a lift and stair core consumes a roughly fixed 150–200 sqft per floor, which is 25–33% of a 600 sqft floorplate but only 8–10% of a 2,000 sqft plate [OWN-CALC]. Small plots lose disproportionately more lettable area at height.
The Land Act (1/2002) then guarantees a permanent supply of exactly the wrong plots. Inheritance subdivision is legal down to 600 sqft (56 sqm) per heir (ss.15, 25); below that, land and building must register as a co-owned building (s.31) [OFFICIAL]. Pre-2002 fragments smaller still exist and trade — the 425 sqft Maafannu listing is one. Combining the height tiers, core arithmetic and the 400 sqft eligibility line, the practical floor for lift-served redevelopment is on the order of 1,000–1,200 sqft, ideally assembled larger [OWN-CALC] — which is precisely what fragmented inheritance makes hard to assemble.
The residual framework, worked honestly
Residual land value is the discipline the quoted prices never face: completed value minus construction, demolition, fees, finance, taxes, contingency and developer profit equals what a builder can rationally pay for the land. Everything in this section is [OWN-CALC] on tiered inputs; nothing is transaction-verified.
The reference project: a 2,000 sqft (186 sqm) mid-block plot carrying a worthless 3-storey walk-up, rebuilt to 8 effective residential storeys — 16,000 sqft (1,486 sqm) gross, 80% efficiency, 12,800 sqft net in 12 units averaging about 1,065 sqft (99 sqm). Completed value is computed two ways: a sale basis (net area × new-build asking prices — legally hypothetical, since Land Act s.19 bars selling or mortgaging buildings separately from land on private Malé plots) and an income basis (capitalised net rent — the legally achievable route).
| Input | Conservative | Base | Favorable | Tier |
|---|---|---|---|---|
| Sale value, completed (MVR/net sqft) | 3,150 | 3,700 | 4,330 | ASKING (D09 new-build band) |
| Rent per unit (MVR/month) | 20,000 | 25,000 | 30,000 | LISTING/OWN-CALC |
| Vacancy + opex (% of gross rent) | 20 | 17.5 | 15 | OWN-CALC |
| Capitalisation rate (%) | 8.5 | 7.5 | 6.5 | OWN-CALC on INDUSTRY-EST yields |
| Build cost (MVR/gross sqft) | 3,200 | 2,600 | 2,000 | OWN-CALC — conflicting ranges, see below |
| Demolition (% of build) | 3 | 2 | 1 | OWN-CALC, no local benchmark |
| Professional fees (% of build) | 8 | 6 | 5 | OWN-CALC, no published scales |
| Contingency (% of hard + soft cost) | 10 | 7.5 | 5 | OWN-CALC |
| Finance (rate on 60% debt; months) | 11.5%; 27 | 11.3%; 24 | 10.0%; 21 | OFFICIAL rate evidence |
| Developer profit (% on cost) | 25 | 20 | 15 | OWN-CALC, no published norm |
| Transaction taxes on land | 0 | 0 | 0 | OFFICIAL — sales tax repealed 2020 |
The build-cost row is the model’s biggest honest uncertainty and a carried conflict between our own dossiers: escalating the Hiyaa official anchor (construction ≈ MVR 1,920/sqft of unit area, 2019 [OWN-CALC on OFFICIAL World Bank cost table]) through +13.53% y/y construction-price inflation [OFFICIAL, MBS] gives MVR 2,500–3,500/sqft for 2026, while contract-anchored evidence in the construction dossier puts mid-market Greater Malé mid-rise at MVR 1,500–2,300/sqft. We run both. The finance rate is the MMA weighted-average lending rate, 11.29% in March 2026 and sticky at 11.3–11.7% since 2020 [OFFICIAL] (MMA via construction-economics dossier). The base build cost of 2,600/sqft blends the escalated-Hiyaa band above with the construction dossier’s contract-anchored band, on a per-unit-area basis.
The base case, line by line:
| Line | MVR m | USD m at peg |
|---|---|---|
| Completed value, sale basis (12,800 sqft × 3,700) | 47.36 | 3.07 |
| less construction (16,000 sqft × 2,600) | 41.60 | 2.70 |
| less demolition (2% of build) | 0.83 | 0.05 |
| less professional fees (6%) | 2.50 | 0.16 |
| less contingency (7.5%) | 3.37 | 0.22 |
| less finance (11.3% on 60% debt, 24 months) | 3.27 | 0.21 |
| less developer profit (20% on cost) | 10.31 | 0.67 |
| Residual land value | −14.53 | −0.94 |
| Residual per sqft of land | −7,263 MVR | −471 USD |
The residual is negative MVR 7,263 per sqft. On the income basis it is worse: 12 units at MVR 25,000/month gross MVR 3.6m a year, MVR 2.97m net, worth MVR 39.6m at a 7.5% cap rate — residual −11,143 MVR/sqft. Across the full scenario grid:
| Scenario | Sale basis (MVR/sqft land) | Income basis (MVR/sqft land) |
|---|---|---|
| Conservative | −21,950 | −28,550 |
| Base | −7,250 | −11,150 |
| Base at D12 contract cost (1,800) | +2,250 | −1,600 |
| Favorable | +6,150 | +6,700 |
| Favorable + 10-storey massing | +7,300 | +7,350 |
| Stretch corner (1,500 cost, 10 storeys, top revenues) | +13,700 | +13,750 |
Read the grid against the evidence ladder. The base case cannot justify any positive land price at all. Only when construction costs drop to the contract-anchored D12 band does land turn positive — around MVR 2,250/sqft on the (legally hypothetical) sale basis. Only the favorable scenario — cheap build, top-of-market rents and prices, thin profit — approaches the bottom of the portal listing band. And only a stretch corner in which everything goes right simultaneously reproduces the administered MVR 15,000. Quoted Malé land values are not redevelopment values. A frontage or corner premium of +10–25% [OWN-CALC assumption] moves none of these conclusions; it is noise against a sign flip.
The model also exposes a replacement-cost inversion worth stating plainly: at base costs, all-in development cost is about MVR 4,030 per net sqft before any developer profit — above nearly the entire MVR 3,150–4,330 new-build asking band [OWN-CALC on ASKING]. At D12 costs it is about MVR 2,790, comfortably inside the band. Whether Malé new-builds are viable at all currently depends on which construction-cost evidence you believe — a conflict only a proper cost survey can settle. The parallel dollar makes it worse than the peg suggests: imported inputs are effectively priced at MVR 20.5-to-the-dollar [MEDIA], which is already inside these MVR cost ranges, while a USD-hurdle investor converting MVR 25,000 rent at the parallel rate receives USD 1,220 a month, not the USD 1,621 the peg implies — a ~25% haircut on USD returns [OWN-CALC on MEDIA rate].
Three break-even inversions summarise the arithmetic [OWN-CALC, base costs, owner-developer basis excluding profit margin]: to earn the broker-midpoint 6.5% gross on cost, a buyer can pay about MVR 1,900/sqft for land; accepting 5.0% gross — deep below the 11.3% cost of money — stretches capacity to about MVR 10,200/sqft; and carrying land at the administered 15,000 requires every unit to rent for about MVR 36,800/month (USD 2,390 at peg) — luxury-outlier territory in a market whose mid-grade 2BR range is MVR 20,000–30,000 [LISTING] and where bottom-60% households must pool 2–4 families to afford MVR 18,900 [OFFICIAL, World Bank].
Why owners neither build nor sell
The yield chart shows the same economics from the landlord’s chair. Buy land at the administered price and redevelop: 4.2% gross on cost. At the listing band: 4.9%. Only inherited land at zero cost basis reaches 6.8% — which is the likely explanation for the broker-quoted 5–8% band, a conflict we report rather than average [INDUSTRY-EST vs OWN-CALC]. Every one of those numbers sits far below the 11.3% lending rate [OFFICIAL]: leverage is negative across the entire capital structure, so debt-financed redevelopment destroys value at asking land prices. Redevelopment pencils only on land you already own — and that is precisely the land the law traps.
Five mechanisms interlock [OWN-CALC synthesis of OFFICIAL and MEDIA evidence]:
- Co-ownership deadlock. Islamic intestacy fixes fractional shares; three generations put dozens of names on one title. Sales execute through the court; transfer beyond the nuclear family requires non-objection of closer heirs (s.17); buyers must be Maldivian citizens (ss.17, 18). One objecting heir blocks sale, mortgage and developer agreement alike [OFFICIAL].
- No strata title on private land. Buildings cannot be sold or mortgaged separately from land (s.19), so there are no pre-sales to finance construction — unlike Colombo or Dhaka. Where strata-like registration exists on new projects it can take up to 4 years, and foreclosure up to 8 [OFFICIAL, World Bank]. Development needs patient equity or lease-share structures.
- The wrong plots. Fragmented inheritance produces exactly the small, awkward plots that lose most to cores and height tiers.
- Holding is free. No recurrent property tax, no vacancy tax, and since 1 January 2020 not even a land-sales tax (the 15% levy was repealed by the Income Tax Act; gains are taxed as income at up to 15%) [OFFICIAL]. The option to wait costs nothing while quoted values ratchet.
- Thin margins at market prices. An owner able to act earns more renting a decrepit walk-up than borrowing at 11.3% to rebuild.
This is a complete causal account of the observable outcome: prime-located, underbuilt, deteriorating stock across all four wards coexisting with five-digit per-sqft quotes and almost no transactions. Sellers quote the option value of scarce land they pay nothing to hold; the only buyers who could rationally pay those quotes cannot exist, because the buildings they would need to sell to fund the price are unsellable by statute. The bid-ask spread is not a market imperfection — it is the law, priced.
What would change the answer
Each lever maps to a specific line of the residual model. A condominium/strata law (planned since the 2019–23 SAP, still absent as of the last authoritative review [OFFICIAL, World Bank]) would legalise the sale-basis column — pre-sales finance, unit collateral, and roughly MVR 3,000–4,000/sqft of additional feasible land value in our grid. A recurrent property or vacancy tax — currently 0% — would, for the first time, make waiting costly and force quoted prices toward residual values; it is the single most price-relevant tax instrument the state does not use. Partition and consent reform (court-supervised sale of co-owned buildings by majority share, rather than unanimity-in-practice) attacks the assembly constraint directly. Construction-cost normalisation — FX unification ending the ~30% parallel-rate import premium [MEDIA] — moves the model from its negative base case toward the D12 cost row, where redevelopment starts to work. Absent these, the rational forecast is more of the same: Malé land will stay quoted, held, inherited — and unbuilt.
What we don’t know
- No transaction prices exist anywhere — the entire price ladder is asking/administered/listing tier; the single “sold” marker is an unverified platform flag; pre-2020 records are contaminated by systematic under-declaration under the old 15% sales tax [MEDIA].
- No ward-level price evidence for Henveiru, Galolhu, Maafannu or Machchangolhi exists in any public source.
- Malé island’s numeric height/FAR schedule could not be retrieved in English; the Hulhumalé 2026 schedule is a benchmark, not the law of Malé [GAP].
- 2026 construction costs are estimate ranges that conflict across our own dossiers (MVR 1,500–2,300 vs 2,500–3,500 per sqft); the residual’s sign flips inside that range. A contractor cost survey is the highest-value missing dataset in this paper.
- No demolition-cost, developer-margin, or frontage/corner premium benchmarks exist; all are labeled assumptions.
- The administered 19-plot sale’s date is unverified (source site unreachable); its 2018–19 vintage is an assessment, not a fact.
- Rental evidence is listings plus a 2021–22 World Bank assumption; no rent index exists, and current HDFC/BML development-lending rates need direct confirmation.