Jumeira Bay Island: a seahorse of plots, and a market of a few dozen properties

What the place is

A small man-made island off the Jumeirah shore, laid out in the shape of a seahorse and joined to the mainland by a causeway. The city centre is minutes away rather than the half-hour the Palm requires, which is the whole geographic argument for it.

The core of the island is the Bulgari resort and residences, with their own beach, marina and yacht club. Around them sit plots for private villas and a small number of houses built to individual designs.

Villa Amalfi is recorded on the island as a 68-unit development by Meraas Holding, completed in 2021 — which gives a sense of the scale involved. Sixty-eight homes would be one building elsewhere in the city; here it is a significant share of the island.

A foreigner takes full freehold and transactions run on ordinary terms. The complication on this island is not the paperwork.

The trap: supply is closed, in both directions

The island is complete and finite in size. No new plots will appear here, which is the rare case in Dubai where supply really is closed rather than merely not yet released.

That is the argument for buying: whatever happens to the wider market, nobody can build a competing island next door, and prices barely sag in a downturn because forced sellers are rare among owners at this level.

It is also the argument for caution, because the same scarcity governs your exit. Few transactions occur in a year, each one is noticed, and finding a buyer when it is your turn to sell can take a very long time.

So the honest way to hold this asset is without a deadline. Anyone who might need to realise the value within two or three years is buying the wrong thing, however good the island is.

Why the usual arithmetic does not work here

Price per square foot on this island is among the highest in Dubai, which tells you where it sits in the market and almost nothing about whether a specific property is well bought.

Yield as a measure means very little. The rental market here is bespoke rather than systematic, and what is being bought is the asset itself rather than an income stream.

There is no depth of comparables to lean on either. A handful of sales a year across a few dozen properties is not a dataset, and a valuation that rests on two transactions should be read as an opinion with two data points behind it.

The useful question replaces all of them: ask how many comparable properties actually sold in the last twelve months, not how many are currently listed. Those two figures diverge sharply in a market this narrow, and only one of them tells you anything.