About the Webb-database RV Dataset
What is this data
The property Rates system (Rates) in HK dates back to the first Rating Ordinance of 1845. Back then, the purpose of Rates was to fund the Police Force. Subsequent Rates were used to fund street lighting (1856), water (1860) and the fire brigade (1875). Since 1931, income from rates has been part of the General Revenue account.
The Rates system relies on the concept of "rateable value" (RV), which is the estimated annual market rental value of a property on a given date. For further context, read David Webb's article.
RV may be used to analyze the General Revenue Account, do macro analysis, cross-check balance sheets and transactions, and research in general.
Is the Data complete?
No. Based on the Rating and Valuation Department's (RVD) Annual Summary, 2024-2025 recorded 2.7m assessments and an Aggregate Rateable Value (ARV) of 744.7bn HKD. Assuming similar values for 2025-2026, we cover 1.3m assessments and an ARV of 392.3bn HKD, which gives us ~53% coverage of ARV.
Due to time constraints we were unable to bring together all assessments. However, one may be able to extrapolate some. If you want to help, please contact us. Extrapolation needs to be done with care, since many assessments group multiple floors, flats or other categories together. We will complement this dataset with completeness statistics and information on what is missing soon. Until then, please be careful when aggregating prices and verify the records.
Data Format
Data was merged with address and location data from the Hong Kong Government Address Lookup Service. Unfortunately, the RVD doesn't have a clear address structure, which makes that harder.
We were able to fully match every building with its coordinates and normalized address. But "building" sometimes refers to estates or multiple grouped buildings with more granular locations. For completeness, we added the full address as specified by the RVD to every assessment.
Beyond trivial normalizations like Rd/Road, Ave/Avenue, St/Street, address numbers are often grouped together and address lines don't always start with the street or include other elements all on one address line. Sometimes street numbers seem to be off by one. Parsing all this comes with several complexities, so we focused on accuracy and didn't want to parse too broadly. We may or may not create a more complex parser in future iterations.
Copyright & disclaimer, Privacy policy
Webb-site.com was originally created by David M Webb MBE
We are committed to maintaining high quality and expanding the database with additional data and insight.