About the House Price Index
- What is the House Price Index?
- How is the HPI calculated?
- Is repossession data included?
- Why is Land Registry's Index different from others on the market?
- How many properties are compared?
- How are the standardised average prices calculated?
- Why is the HPI revised each month?
- Is the HPI adjusted for seasonality?
- Is the Land Registry HPI adjusted for inflation?
- What regions are covered by the HPI?
- What do the reported 'monthly' and 'annual' price changes mean?
- What data is used in the Land Registry HPI?
- Can the data be re-published?
- Why are small local authority districts and postcode sectors not shown on the HPI?
- Why has Land Registry decided to release repossession data?
- Why does the headline figure reported by Land Registry continue to exclude repossessions?
- Where does the repossessions data come from?
- What repossessions data does Land Registry publish?
- What is the impact of including or excluding repossessions data from the HPI?
- Is the repossessions-adjusted HPI seasonally adjusted?
- Why are volumes for the last few months not reported?
- How much repossessions data is there?
- Why does Land Registry's data differ from other available sources of repossession information?
- Will repossession data presented in the House Price Index background tables be published separately?
- Why are you releasing this data?
- How does this data differ from that available from the ‘Search the House Price Index’ facility?
- Will you publish the background data for commercial properties and land only sales?
- What format will the data be in?
- Where can I assess the tables?
- What is the difference between the tables called Indices, Indices (SA) and Indices (SA & SM)?
- What is the difference between the tables called Average Prices (SA) and Average Prices (SA+SM)?
- What data is presented in the Monthly Change and Annual Change tables?
- What is to be found in ‘Sales Volumes’ table?
- To what level is the data presented?
- Why do some counties have blank values?
The Land Registry House Price Index (HPI) captures changes in the value of residential properties.
The HPI is published by Land Registry using sales data collected on all residential housing transactions, whether for cash or with a mortgage, in England and Wales since January 1995. You can search indices at a national level as well as for the various regions, counties and London boroughs.
Our HPI uses a sample size that is larger than all other statistical measures available. The Land Registry data set is the only complete record of residential property transactions in England and Wales.
The HPI is calculated by using Land Registry's own 'Price Paid Dataset'. This is a record of all residential property transactions made in England and Wales since January 1995.
At present it contains details on over 16 million sales. Of these, over six million are identifiable matched pairs, providing the basis for the repeat-sales regression analysis used to compile the index. This technique of quality adjustment ensures an 'apples to apples' comparison between properties.
The HPI is a repeat sales regression (RSR) index, measuring average price changes in repeat sales on the same properties, ensuring a like for like comparison. This means that price changes on a flat in Mayfair are not compared to those on a flat in the Old Kent Road.
The statistical computation of the HPI is performed by Calnea Analytics Limited. Full details of methodology, a discussion of technical questions and a comparison with other index creation methods can be found by visiting www.calnea.com
From 29 January 2013 the House Price Index PDF monthly report will contain two additional pages which provide information on repossessions recorded at Land Registry. For more information on this data see our reposessions FAQ.
Land Registry's House Price Index uses a sample size that is larger than all other statistical measures. The Land Registry data set is the only complete record of residential property transactions in England and Wales. In addition, Land Registry's House Price Index is created using the Repeat Sales Regression method. This technique of quality adjustment ensures an 'apples to apples' comparison between properties.
The HPI is calculated by using Land Registry's own 'Price Paid Dataset'. This is a record of all residential property transactions made in England and Wales since January 1995. At present it contains details on over 16 million sales. Of these, just over six million are identifiable matched pairs, providing the basis for the repeat-sales regression analysis used to compile the index.
The standard average house price presented by Land Registry is calculated by taking the average (geometric mean) price in April 2000 and then recalculating it in accordance with the index change both back to 1995 and forward to the present day.
Historical data published as part of the HPI are revised primarily for two reasons:
Data capture lag - There is a time-lag between the sale of a property and the subsequent registration of this information with Land Registry. This means the transaction data for the previous month will be incomplete when the monthly report is prepared. The missing data is included as it becomes available.
HPI based on repeat transactions - When Land Registry publishes its HPI reports each month, the index represents the best available view on historic house price movements at the time. As new information becomes available, the published indices are revised to reflect any new data. It is seen as statistical best practice for time-series indices to restate historic data as new information becomes available.
Yes, the Land Registry HPI is designed to remove the influence of seasonal variations on house price changes.
The HPI measures nominal house price changes and is not adjusted for inflation. Users of the HPI can make their own inflation assumptions and adjust the HPI accordingly.
The index provides information for each of the ten Government Office Regions (GORs) of England and Wales based on the latest definitions from the Office of National Statistics. At present these are:
- North East
- North West
- Yorkshire and The Humber
- East Midlands
- West Midlands
- South East
- South West
The 'annual' percentage change in house prices is simply the price change relative to the same month twelve months earlier. For example, if the HPI release is for February, then the 'annual' price change reports the percentage change in values relative to February in the previous year.
This annual rate of price change reflects the best estimate for how much the value of a typical property has changed over the preceding twelve months. Similar to the 'annual' change, the 'monthly' percentage change estimates the percentage change in home values relative to the previous month.
The Land Registry HPI is derived from all residential property transactions registered with Land Registry since January 1995. See the list of exclusions for further information.
Yes, but not for commercial purposes. No part of the HPI may be reproduced in any form or for any purpose without the prior permission of Land Registry. Enquiries can be made to copyright@landr egistry.gov.uk. Neither Land Registry nor any third party shall be liable for any loss or damage, direct, indirect or consequential, arising from (i) any inaccuracy or incompleteness of the data in the HPI or (ii) any decision made or action taken in reliance upon the data. Neither shall Land Registry or any third party be liable for loss of business resources, lost profits or any punitive indirect, consequential, special or similar damages whatsoever, whether in contract or tort or otherwise, even if advised of the possibility of such damages being incurred.
Due to the small volume of sales for these areas, the HPI would not always provide a reliable measure of price movement. This may become possible as the amount of data increases.
As a member of the Public Data Group and as part of a wider Government initiative, we are committed to releasing as much of our data as possible.
In 2012 we received feedback from a number of customers showing a significant interest in repossession sales. This prompted us to investigate issues such as what we include in our HPI data and its impact on our customers and any wider audience. As a result of our findings we decided to publish additional information about repossessions.
There are no changes to the existing headline statistics or any other element of the Land Registry HPI report as the majority of our customers have expressed a preference for the data to remain in its current format. However, we wanted to include additional information about repossessions for those customers that are interested in this option.
To ensure consistency with our previous reports we will continue to report the headline data excluding repossessions. The data used to calculate the repossession statistics is extracted from a different source and so may contain other transactions currently excluded from the HPI, such as potential commercial properties.
For these reasons and due to the data only being available from 2006 onwards, we are unable to include repossession transactions in the main HPI.
Land Registry receives transfers by lenders exercising their power of sale. Once we have identified these transactions, we are then able to extract the Price Paid Information from the related register entry as opposed to the price paid dataset used to collate Land Registry’s Market Trend Data.
Price Paid has been recorded on the register since 1 April 2000 and entries recorded from January 2006 onwards are included in the repossession information.
From 29 January 2013 the HPI report incorporates repossession volumes for England and Wales going back to 2006, broken down by region for the current month together with an adjusted headline statistic which includes the additional repossession data.
Research shows that repossessed properties tend to sell quicker and achieve a lower price than equivalent properties not sold under a lender’s power of sale. During a period of market decline, including repossessions in the overall HPI tends to increase the reported price falls. Conversely when the market rises, including properties that were originally purchased as repossession but have been resold tends to increase the reported price rises.
It is important to bear in mind that repossessions only represent a small portion of total transactions. This is the main reason that the impact on the HPI is minimal.
While the HPI goes back to 1995, the repossessions data was only comprehensively recorded after 2006. It is for this reason that prior to 2006, the HPIs with and without repossessions data are identical.
Yes, we have used identical seasonal adjustment factors for both versions of the HPI.
Land Registry has partial sales volume data for the months immediately preceding the HPI release. However this is not published precisely because the data is not complete. We aim to report on accurate and reliable information and therefore do not want to release sales volume data that has the potential to be misleading if quoted out of context.
Between January 2006 and December 2012, we have data on approximately 124,000 repossession transactions.
The repossession data used by Land Registry in the monthly HPI report is based on transactions lodged with us by lenders exercising their power of sale. The table below shows some of the differences between the different suppliers of this information:
|Supplier of data||Coverage||Regional stats available?||Frequency||Seasonal adjustment ||Publication lag||Reports||Exclusions||Additional information|
All transactions under a power of sale lodged with Land Registry or Actual possessions
Primarily non-residential properties (please see our full FAQs for more detail)
Dependant on lodgement at Land Registry
Council of Mortgage
1.5 - 2 months
All 1st Charge mortgages taken into possession from CML Members only.
2nd Charge mortgages, commercial and bridging loans
Reports based on CML data (as above)
1.5 - 2 months
All County Court possession actions (claims issued and orders made)
Voluntary and Actual repossessions, since not all possession orders are executed.
Mortgage Accounts where property is in possession. (voluntary and court-ordered)
If more than one charge on property possession may be recorded twice.
No. While the HPI goes back to 1995, the repossession data was only comprehensively recorded after 2006. Due to this and the sensitive nature of this data only the most recent high level aggregated information is available for statistical purposes via the PDF. It is important to bear in mind that repossessions only represent a small portion of total transactions. This is the main reason that the impact on the HPI is minimal.
We are committed to releasing more of our data as part of our commitment to the Open Data agenda. We already publish the House Price Index and Price Paid Data, which help to support business innovation and stimulate the economy. Through greater openness we hope to encourage more creative individuals and organisations to develop innovative applications and services that benefit industry and the public.
The Search the House Price Index facility available on our website is for those customers that would like to produce printable reports derived from the background House Price Index data. Searches are available covering average prices, indices, sales volumes and by property type and ranges from January 1995 to the most recent available data.
The House Price Index background tables are for our customers that would like to download the full dataset in a more usable format which incorporates both the raw and cleansed aggregated data behind the House Price Index. The tables are made up of:
- Indices Seasonally Adjusted (SA)
- Indices Seasonally Adjusted and Smoothed (SA & SM)
- Average Prices Seasonally Adjusted (SA)
- Average Prices Seasonally Adjusted and Smoothed (SA & SM)
- Monthly Change
- Annual Change
If you would like to understand more about the tables above and the differences between the figures, please read the Technical information below.
No as we do not currently hold this data. The House Price Index is derived from our Price Paid Data, which only reports on single residential properties sold for full market value. Please see Price Paid Data exclusions for more information.
What format will the data be in?
The data will be available in Excel and CSV format initially and then in linked data format from July 2013.
The tables will be published monthly in Excel and CSV format and will be available from our website and Data.gov.uk
- The Indices table contains raw monthly residential property price indices at Regional, County/Unitary Authority and London Borough level.
- The Indices (SA) table contains seasonally adjusted monthly residential property price indices at Regional, County/Unitary Authority and London Borough level.
- Indices (SA+SM) table contains smoothed seasonally adjusted monthly residential property price indices. The Regional HPI data is not smoothed. Smoothing only applies to County/Unitary Authority and London Borough level. The smoothing is done by calculating a rolling four-monthly (transaction weighted) average. These figures are representative of those reported in both the PDF doc and Search Facility.
- The Average Prices (SA) table contains seasonally adjusted standardised average prices at Regional, County/Unitary Authority and London Borough level. These are derived through the application of the seasonally adjusted price index to the geometric mean price for each area at the start of the time series.
- The Average Prices (SA+SM) table contains smoothed seasonally adjusted standardised average prices. These are derived from seasonally adjusted standardised average prices by applying rolling four-monthly (transaction weighted) averages. These figures are representative of those reported in both the PDF doc and Search Facility.
- The Monthly Change table contains monthly change figures which represent the month-to-month percentage change of the index. Data is provided down to a regional level only.
- The Annual Change table contains annual change figures. These represent the percentage change in the index over the last 12 months. This data is provided down to a regional level only.
This table contains counts of all sales included in the electronic register at the time of data extraction. The data collection lag time predominantly affects the most recent months, therefore the most recent two months are omitted from this data to avoid reporting misleading volumes.
Data in ’Indices‘, ’Indices (SA)’, ’Indices (SA+SM)’, ’Average Prices (SA)’, ’Average Prices (SA+SM)’ and ’Sales Volumes‘ are provided down to Regional, County/Unitary Authority and London Borough level. Please note that smoothing is not applied to regional level data. Data in ’Monthly Change‘ and ’Annual Change‘ tables are presented at a regional level.
A ’not applicable’ or ‘blank’ value will be shown for a county or region if there were no sales that meet the HPI inclusion criteria within the area for that month.
The index is calculated using repeat-sales regression. County/region calculations are based on registered properties within the county which have been sold for full market value since 1995.