Financial Times FTHPI House Price IndexThe FT House Price Index (FTHPI-MA) is calculated monthly by Acadametrics based on the (approx) 120k monthly residential property transactions reported to the Land Registry (LR) for England and Wales. This gives the FTHPI-MA an almost complete sample, and the advantage of including the (approx) 25% of property that is bought with cash rather than a mortgage, and useful because cash purchases sometimes follow trends different form those of mortgaged purchases. TimelinessHowever, property transactions can be registered up to two years after completion, and so the Land Registry data, while being authoritative, suffers from being both delayed and incomplete, particularly for the most recent months. The timeliness required of an index, especially if it to reflect what is currently happening in the market, is provided in the FT House Price Index by forecasting the latest month based on an "optimal" portfolio of indicators that are chosen to minimise bias and variance from the (eventual complete) LR record. The index is updated and revised as and when the LR transaction data becomes available. For more details, see this summary table: A Comparison of UK House Price Indices.
Figures 1 & 2.
House prices and annual house price inflation; HPI curves are on a corrected timeline with YoY shifted by 6 months and QoQ by 6 weeks.
Mix adjustmentsThe original FT House Price Index index was based on a simple average of the LR sold price series, but since July 2005 the index has been mix-adjusted to take into account variations in the mix and region of houses sold. The new series appears to be calibrated to reflect the mean price of the transactions - different from the Halifax and Nationwide that aim to reflect the price of a typical house and more closely match the LR median price. The mean price is usually higher than the median price because the typical distribution of house prices is significantly skewed with a long upper tail. But note that the LR has moved to a new index, the LR-HPI, based on a repeat sales regression and calibrated to reflect a typical house valued close the the median transacted price.
VolatilityOne of the aims of the FT House Price Index was that it should provide monthly information on house prices that did not suffer from the volatility inherent in the smaller samples of mortgage approval data used by the Halifax and Nationwide; this is clearly achieved by using the full Land Registry data, but at the expense of timeliness, and moving back to the smaller sample indicators for the forecast element reintroduces some volatility; this is countered by additionally smoothing the index over a three month window. Seasonal adjustmentsThe level of prices and transactions in the LR record show significant variation that reflects the seasonal changes in house buying activity over the course of a year; the FT House Price Index also provides a seasonal correction to the LR record. ResourcesAcademetrics FT House Price Index |
See Also ... |
