UK House Price Indices Suffer from Incomplete Sales Data

by | Feb 18, 2022 | Insights, Property Analytics

  The Land Registry Price Paid Dataset provides a complete database of prices at which residential properties are transacted in England and Wales. Data is released with a two-month lag so that a dataset published on the nth month will contain transactions that took place up to month n-2.

Not only is there a lag to contend with but the datasets that are released are incomplete. For example, the data set released in February 2020 only included transactions up to the end of December 2019 and contained only between 20% to 30% of all the monthly property transactions reported by the HM Revenue and Customs (HMRC) that occurred in December 2019. The March 2020 data set included between 60% and 70% of all December 2019 transactions and the April 2020 data set contained between 80% and 90% of all transactions that occurred in December 2019.

If we look at transactions that occurred after the beginning of the pandemic then we see that the data emerges even slower than before the pandemic. For example, the data set released in December 2021 contains between 3% to 5% of transactions for October 2021. In the January 2022 dataset, we see that between 20% to 30% of October transactions are recorded and in February between 30% to 40% of transactions are recorded.

Figure 1 shows MIAC’s estimate of the average Land Registry data emergence over a 12-month period in 2019 and post-pandemic from April 2020. The impact of the pandemic on data emergence is clear from the difference between the blue and orange lines and we can see how the most recent 12-month period has not shown any improvement.

 

Figure 1. Land Registry Data Emergence

A number of house price indices in the UK utilise Land Registry data since it is the only database of actual transactions that occur within the UK and Wales. Analysing a number of freely available indices that rely on Land Registry data shows that the most recent months have much greater volatility than prior months. The source of this volatility is not the housing market but the lack of data caused by the slow emergence of Land Registry data which has worsened since the pandemic.

Figure 2 shows the regional ONS indices and their rolling 6-month volatility. The surge in volatility for the most recent data points can be clearly seen in the graph.

Figure 2. : Regional ONS Indices & their volatilities

Indexing property portfolios to the most recent data using indices where the most recent data is the most volatile and the one that is most likely to be restated in subsequent months is therefore subjected to a significant amount of error. This error is also likely to compound if all properties are indexed from month to month rather than fromthe original valuation.

MIAC utilise more real time housing data and use this to supplement the Land Registry data in order to remove this volatility and provide the most accurate up to date house price index with the lowest lag possible.