Dubai offices against London and New York: rents, yields and what the tax does

Comparing office markets by headline rent is pointless: what reaches the owner is decided by the form of ownership and by tax. So all three parameters at once, across three global business centres.

London, City of London

Prime office rent of roughly AED 9 000 per square foot a year and an average yield near 5.25%. Ownership is predominantly leasehold — a right for a term rather than ownership. Taxes: income, capital gains, stamp duty, VAT and inheritance.

A high entry threshold against a modest net yield. Leasehold adds a problem of its own: the unexpired term affects resale value more than the condition of the building does.

New York, Manhattan

Class A offices at AED 7 000–8 000 per square foot a year, yields of 3–4%. Ownership is freehold. Taxes: income, property, capital gains, transfer on purchase, and estate tax.

Ownership is complete, but the tax load consumes a substantial share of the gross yield, and annual property tax is paid whether the building is let or standing empty.

Dubai, Business Bay

Average rent around AED 450–490 per square foot a year in that cluster, with class A across the city averaging AED 370 after a 28% annual rise. Ownership is freehold in designated areas. For an individual there is no property tax and no capital gains tax, and ownership above the threshold supports a residence visa.

The rate per foot is a fraction of London's or New York's, and so is the entry price. What survives the comparison is the net figure — the absence of an annual holding tax and of capital gains tax changes the arithmetic more than the rent differential does.

What the comparison is actually for

Not to declare a winner. London and New York are deep, liquid, institutionally traded markets where an owner accepts a lower net yield in exchange for that depth. Dubai offers a higher net figure with a shallower buyer pool and a shorter history to price against.

The useful takeaway is methodological: when comparing any two office markets, compute the number after tax and after the ownership form, not the rent. The ranking changes.