The OECD claims that UK house prices “significantly exceed long-term averages relative to rents”. Really?

The OECD calculation compares the ONS house price index with the rents component of the RPI. These measures, however, are based on different “baskets” of houses. The OECD is comparing apples with pears.

Specifically, the ONS house price index is a value-weighted measure (i.e. giving greater weight to expensive houses) based on a survey of mortgage-financed transactions. It excludes public sector housing.

The RPI rents index, meanwhile, measures actual payments by tenants in both the private and public sectors. It is implicitly weighted towards lower value housing. Public sector rents (i.e. paid to housing associations and local authorities) account for nearly two-fifths of the index.

A far superior methodology is to compare national accounts data on aggregate rents – both actual and imputed to owner-occupiers – with the value of the housing stock. Assuming that the statisticians use appropriate methods for imputing owner-occupied rents, the two parts of the ratio will reflect the same mix of housing.

This national rental yield measure is regularly calculated here and signalled that house prices were about 30% overvalued at their peak in 2007 but had moved to significant undervaluation by end-2011. This contrasts with the OECD's claim that the price to rent ratio has remained above average in recent years.
 
What is the current message? The national rental yield was an estimated 4.52%* at end-2013 versus an average since 1965 of 4.26%, implying 6% undervaluation. Recent house price strength, in other words, has not been excessive relative to housing market “fundamentals”, which have simultaneously driven up rents.

Houses are expensive in terms of earnings but bubbles are characterised additionally by an overshoot of prices relative to rents, as rose-tinted perceptions of capital gains potential distort assessments of the merits of owning rather than renting. Current prices are still far from bubble territory, though could get there if monetary conditions remain excessively loose.

*The yield is calculated on a trailing 12-month basis, e.g. its end-2013 value equals rents in calendar 2013 divided by the December housing stock . The latter was estimated by linking the published end-2012 value to the change in the ONS house price index between December 2012 and December 2013.