Pandemic Leading to Longer Time Lags in Registration
Creates Problems With Short-Term Data & House Price Indices
There has always been a time lag in when transactions are registered with HM Land Registry but the pandemic has increased it. The increased lag creates a short-term constraint on our understanding of the housing market with less comprehensive transaction data and could lead to larger revisions in the ONS house price index.
There is a time lag in when transactions are registered with HM Land Registry and the pandemic has made it worse. The lag in registrations means that the latest data release from Land Registry only covers a fraction of transactions that actually occurred in the latest month it covers. This lag and partial sample of transactions has caused problems in the past. For example, the revision period of the ONS house price index was extended from three to twelve months and numerous people have mis-interpreted the limited number of transactions in the Price Paid Data as a massive collapse in housing market activity when it was just the lag. Normally we can make allowances for the lag but, unfortunately, the pandemic has worsened it and this has implications for the data.
The chart opposite shows the impact of the lag on sales in the Land Registry’s Price Paid Data. It shows limited sales in the most recent month’s release slowly increasing towards the number recorded by the HMRC with each release. For example the first count of sales in Sep 2020 was just 880 and had increased to 62k by the Dec 2020 release. However, the HMRC data reports nearly 86k sales in Sep and so the Land Reg data will continue to close towards this level over time, though it never quite reaches the same level due to differences in data coverage
Unfortunately, it appears that a combination of the pandemic and rising transaction numbers have worsened the lag in recent months. The increasing lag can be seen in Figure 2 below which looks at what proportion of sales (based on HMRC data) have been registered for the month by the number of data releases covering that month. For example, it shows 43% of HMRC sales in July 2020 had been registered by the second release of data covering that month but the second release for November covered just 20% of sales. Meanwhile, a comparison with data for 2016 shows that we would’ve expected around 73% of sales to appear in the second release.
Hopefully this increased lag is just a short-term effect reflecting the challenges faced by Land Registry during the pandemic and a boom in transactions. There are, however, some serious consequences while it lasts. The increased lag means that the number of sales in the third release for any given month will still be a substantial undercount (this is when ONS release sales data). It also implies the first ONS house price index for any given month is based on a very small sample of sales. We could therefore see some significant revisions to the index, especially at smaller geographies.