Methodology

The rules every market follows

• One population per question. Registers that record different events are never pooled into one figure.
• Every count carries its window. A number with no stated period is not published.
• The count beside a figure is that figure’s own denominator, never a larger number borrowed from a wider sample.
• Thin samples are named. Below a market’s sample floor, thirty transactions in Singapore, the figure is withheld with its reason in the open, not quietly rounded into existence.
• Every published figure is reproducible from a static endpoint that market serves.
• No valuations, no forecasts, no advice. These are statistics about transactions that already happened.

Market by market

Why the two markets are not one table

The differences below are not rounding: they are why IndexProp publishes each market on its own terms rather than ranking them against one another.

• Different populations. Singapore’s national figure is HDB public housing, sold on 99-year leases under eligibility rules. The United Kingdom’s is the whole private market.
• Different floor-area provenance. One market’s denominator is an Energy Performance Certificate’s measured area; the other’s is the housing authority’s own field on the same record as the price.
• Different mix adjustments. The United Kingdom weights property types by their share of sales in a fixed 2013 to 2017 base window and chain-links each period to the next; Singapore weights flat types by the national dwelling stock, because a transaction weight would carry the very mix bias the adjustment removes. Both are defensible and they are not interchangeable.
• Different anchors. The two markets’ twelve-month windows end in different months, because each register is anchored on the last month its own filings have substantially arrived.
• No exchange rate. Neither market ingests one. A converted price gap would move on days when no home changed hands in either country.

What IndexProp does not publish

• Cross-market price comparisons. Deferred until the obstacles above can be stated in the figure rather than around it.
• Rental-yield comparisons. One market publishes yields and the other publishes no rent data at all.
• Affordability against income. One market publishes earnings; the other publishes none at a matching geography.
• Housing-stock comparisons. One market publishes dwelling counts; the other publishes a sales-share weight, which is a flow and not a stock.
• Any pooled figure. A median across two countries’ registers would describe neither.

A third market, a rent register on the UK side, or an income series at a matching Singapore geography would each make part of this answerable. When one lands, the comparison it supports ships with its caveats attached, not retrofitted onto a headline that already exists.

Sources and licences

Every figure belongs to the market that published it and carries that market’s licence. Singapore’s HDB-derived figures are under the Singapore Open Data Licence version 1.0; its private figures are derived statistics from the URA Data Service, whose records are not redistributed. The United Kingdom’s figures contain HM Land Registry data under the Open Government Licence v3.0. Each market’s own pages state the split in full.