Building land out of the sea

The question nobody asks in the right order

Maldivian land policy has had one answer for thirty-five years: when a place runs out of room, make more of it. Hulhumalé I in 1997, Phase II in 2015, Phase III from 2023, Gulhifalhu, Thilafushi, Addu, and now RasMalé at up to 1,153 hectares.

The debate about this is almost always environmental or political. The economic question is asked far less often, and it is the one that determines whether any of it works: what does a square foot of created land cost, what does it sell for, and who captures the difference?

The Maldives publishes no answer. There is no cost breakdown for any reclamation project, no land-use split showing how much of a reclaimed area becomes saleable plot, and no independent audit of progress on anything. What does exist is two disclosed dredging contracts and a set of official land prices. Putting those two things next to each other is the whole of this paper, and it is enough to change how the market should be read.

What making land costs

Two projects have published enough to compute a rate:

• Gulhifalhu: roughly USD 103m of Boskalis dredging contracts over 2020–2024 for about 150 ha [MEDIA on OFFICIAL] — MVR 98 per gross square foot at the peg.
• Addu City: an India EXIM-financed package of about USD 80m for roughly 190 ha including three tourism islets [OFFICIAL] — MVR 60 per gross square foot.

Not all reclaimed area becomes saleable. Roads, open space and utility corridors take a substantial share; no Maldivian project publishes its split, so we assume 50–60% saleable [OWN-CALC, stated assumption]. That gives MVR 101–197 per saleable square foot as the cost of land formation.

This is much cheaper than the discourse implies. Reclamation is not the expensive part of Maldivian development. Construction is: our applied feasibility analysis puts building cost at roughly MVR 2,500 per saleable square foot, more than twelve times the cost of the ground it stands on.

What land sells for

Against MVR 101–197 of creation cost:

• Thilafushi Phase 2 industrial land was offered freehold at a base of MVR 1,700 per square foot [OFFICIAL — HDC and Gazette, May 2024]. That is 9–17 times cost.
• Hulhumalé Phase 2 residential plots carry auction floors of MVR 4,000–4,500 per square foot [OFFICIAL]. That is 20–45 times cost.

The residual at the Hulhumalé floor is roughly MVR 3,800–3,900 per saleable square foot. That number is the subject of this paper, because it is entirely unexplained in public.

It contains some genuine costs. Trunk utilities — water, sewerage, power distribution, roads — are real and expensive, and D14 of our real-estate programme records that Gulhifalhu’s land value is explicitly gated on trunk utilities that remain unfinanced. Coastal protection is real. So is the financing cost of holding reclaimed land for years before it generates revenue, at a sovereign that pays distressed-sovereign yields.

But it may also contain economic rent — a monopoly seller pricing scarce land at what the market will bear rather than at cost recovery. Both explanations fit the observable data equally well, and no published Maldivian document distinguishes them. That is not a rhetorical point. It is the single most consequential unpublished number in Maldivian development economics, because it determines whether plot prices can fall.

The state’s two valuations of the same land

There is one place where the state has valued reclaimed land explicitly, and it does not agree with its own auctions.

The Auditor General assessed HDC’s loss on the Binveriya free-plot scheme — 1,407 Phase 2 plots, 2.84m square feet — at MVR 14.85–15.93bn [OFFICIAL]. Per square foot that is MVR 5,229–5,609, which is 16–40% above the MVR 4,000–4,500 auction floors for comparable land.

Two official numbers, same land, materially different. Either the auction floors understate market value, or the audit valuation overstates what the plots would clear. We carry both rather than averaging them, but anyone pricing Hulhumalé land should know that the state’s own two assessments differ by up to two-fifths.

The Binveriya episode matters for a second reason. Giving away 1,407 plots destroyed the land-sales revenue that funds HDC’s infrastructure programme — and HDC is simultaneously carrying MVR 628m of Hiyaa tenant arrears with a 37% rent-to-own default rate [OFFICIAL/MEDIA]. The entity that creates most of the country’s new land is under financial stress caused by giving that land away, which is a supply risk rather than a governance footnote.

Does reclamation pay?

Set out as a model rather than a slogan:

  LAND CREATION COST                    per saleable sqft
    dredging & formation                MVR 101 – 197      [evidenced]
    coastal protection                  not published      [gap]
    trunk utilities                     not published      [gap]
    roads & public realm                not published      [gap]
    financing over the holding period   not published      [gap]
  ────────────────────────────────────────────────────────
  LAND VALUE REALISED
    industrial (Thilafushi floor)       MVR 1,700          [official]
    residential (Hulhumalé floor)       MVR 4,000 – 4,500  [official]
    state's own valuation (Binveriya)   MVR 5,229 – 5,609  [official]
  ────────────────────────────────────────────────────────
  RESIDUAL                              MVR ~3,800 – 3,900
                                        composition unknown

Four of the six cost lines are unpublished. So the honest answer to “does reclamation pay?” is: on the disclosed evidence it appears to pay by a very wide margin, and that margin is wide enough that it would survive substantial unpublished costs — but nobody can demonstrate it, including the state.

What we can say with more confidence is the comparative claim. Reclamation is cheap relative to construction, relative to land prices, and relative to the infrastructure that has to follow it. The binding constraint on Maldivian land supply is not the dredger. It is the balance sheet that has to fund everything after the dredger leaves.

RasMalé, and the limits of announcement

RasMalé is the test case for that constraint.

At the rates the two disclosed contracts imply, reclaiming 1,009–1,153 ha costs USD 425–792m [OWN-CALC]. The only cost ever published for the project is a USD 21m termination payment to a contractor [MEDIA]. Usable reserves in April 2026 were USD 148–244m [OFFICIAL].

The dredging alone, in other words, plausibly costs two to five times the country’s entire usable foreign reserve position — before a single metre of water main, road or building. The announced 65,000 housing units have no identified financing at all.

Our assessment, consistent with the real-estate programme’s: reclamation will probably finish, because roughly 60% was already done by June 2026 and sunk projects tend to be completed. The city on top of it should be assigned a 5–25% probability of delivery at announced scale within ten years [OWN-CALC]. For anyone modelling Greater Malé supply, RasMalé is a sentiment effect on today’s prices, not future stock.

What actually completes

The 2025 record is instructive because two projects did complete — and both were foreign-financed. Velana International Airport’s Terminal 1 opened in July 2025, six years late, at roughly USD 585m against a USD 350m award [OFFICIAL/MEDIA]. Hanimaadhoo opened in November 2025 on a USD 136.6m India EXIM line of credit [OFFICIAL].

That is the pattern. Maldivian infrastructure completes when a foreign balance sheet carries it, and slips or stalls when it depends on domestic fiscal capacity. Given the position set out in How the Maldivian economy actually works — debt at 129.7% of GDP, high assessed risk of distress, nine to fourteen days of usable import cover — that is unlikely to change this decade.

The practical consequence for property underwriting: the probability of delivery belongs in the model, not in the narrative. A land-value uplift thesis resting on the Thilamalé bridge is materially different from one resting on RasMalé, and the difference is roughly 92% against 15%.

Who captures the value

Three groups, in descending order of certainty:

1. Existing landholders in the corridor. Uplift from connectivity accrues to whoever already owns nearby land, without their doing anything. The Thilamalé bridge re-rates Villimalé and activates Gulhifalhu and Thilafushi; the beneficiaries are current holders, not the financier.
2. The state, as monopoly seller of new land — but only if it sells rather than allocates. Binveriya is the counter-example, and it moved roughly MVR 15bn of value from the balance sheet that funds infrastructure to individual recipients.
3. Developers, last and least. Our feasibility analysis finds that market-rate residential development does not clear its cost of capital at auction-floor land prices. The value created by reclamation is captured in the land price before a developer sees it.

The Maldives has no land-value-capture mechanism — no betterment levy, no infrastructure charge, no property tax, and no transfer tax outside special economic zones [OFFICIAL]. Uplift created by publicly-financed infrastructure accrues privately and is never recycled into the next project. Given that the binding constraint on land supply is the balance sheet, that is the most consequential policy gap in the sector.

What we do not know

• No Maldivian reclamation project publishes a cost breakdown. The two rates in this paper come from disclosed contract values divided by disclosed areas, and cover dredging only.
• No project publishes a land-use split, so the saleable share is a Nyra assumption of 50–60% rather than a measurement.
• Trunk utility, coastal protection and financing costs are unpublished for every project, which is why the residual between creation cost and sale price cannot be decomposed.
• Auction floors are minimum bids, not clearing prices. Winning premia are unpublished, so the realised value of Hulhumalé and Thilafushi land is unknown.
• The Auditor General’s Binveriya valuation and HDC’s auction floors disagree by 16–40% for comparable land, and we cannot reconcile them.
• No independent engineering audit exists for any project. All progress percentages are government or contractor statements reported through the press.
• RasMalé’s masterplan has never been published in technical form — no land use, no phasing, no utility budget. The 65,000-unit figure is an announcement.
• Delivery probabilities in this paper are structured judgements, not statistics, and are published so they can be argued with rather than because they are measured.
• No cost-benefit or land-value-uplift study has ever been published for a Maldivian infrastructure project, so every uplift claim in this sector — including ours — is inference.

Sources

• Boskalis / Gulhifalhu dredging contract reporting (2020–2024)
• India EXIM line-of-credit disclosures — Addu reclamation, Hanimaadhoo airport
• HDC — Hulhumalé Phase 2 plot sales and Thilafushi Phase 2 freehold sale (Government Gazette, May 2024)
• Auditor General — HDC Binveriya scheme audit (via edition.mv)
• ADB project 51077-003 and AIIB financing documents — Greater Malé waste-to-energy
• MACL — Velana International Airport Terminal 1
• Construction Ministry statements via Atoll Times, Sun, Avas, PSM, The Edition
• IMF — 2026 Article IV mission statement; World Bank — Maldives Development Update, June 2026
• Nyra — real-estate programme dossiers D04, D07, D08, D14