Discussions about how far house prices could fall often confuse two issues – the extent of current overvaluation and the possibility that a serious economic downturn will result in prices undershooting “fair value”. This post focuses on the former.

According to widely-quoted IMF research (here, p.11), UK prices rose by nearly 30% more than justified by “fundamentals” between 1997 and 2007. Some commentators – including the MPC’s David Blanchflower (here, p.5) – have used this 30% figure as a measure of current overvaluation. As David Smith of the Sunday Times pointed out in a recent column, this is incorrect because the reference is the level of prices in 1997, not “fair value” at present. Adjusting for price growth since 1997, the IMF calculations imply 15% overvaluation currently.

The IMF approach may be questioned. The definition of “fundamentals” includes affordability, income growth, interest rates, credit growth, equity prices and population trends. The relevance of credit expansion and equity prices for sustainable valuation is debatable. In addition, no allowance is made for constraints on the supply of housing – likely to have been a more significant factor in the UK than in other countries recently experiencing house price booms.

Popular comparisons of the house price to earnings ratio with its long-run average significantly overstate current overvaluation. The first chart shows one such measure – the value of the housing stock divided by household disposable income. The ratio clearly trends higher over time, reflecting factors such as improving quality, the pressure of an expanding population on constrained supply and a high income elasticity of demand for housing.

Rather than its long-run average, the log-linear trend of the ratio is probably a better guide to current “fair value”. On this basis, prices were 10% overvalued at the end of last year versus an 85% deviation relative to the average.

An alternative and superior approach is to use rents rather than earnings as the basis of comparison. Rents already embody fundamental influences on housing demand and supply. Moreover, a potential homebuyer does not face a choice between consuming housing services or retaining earnings for other purposes – the decision is rather whether buying is financially preferable to renting.

The second chart shows a measure of the rental yield on housing derived from the national accounts – actual and imputed owner-occupied rents as a percentage of the value of the housing stock. The yield stood at 3.0% at the end of 2007 against a long-run average of 3.6%, suggesting price overvaluation of 20%.

However, this figure overstates the need for prices to fall to restore value, for two reasons. First, rents are growing solidly – by 7.7% on the national accounts measure in the year to the fourth quarter. The trend seems likely to persist, with the latest RICS letting agents’ survey reporting strong tenant demand and expectations for rents – see third chart. Secondly, low real yields on competing assets – particularly government bonds – may mean the “equilibrium” rental yield is below its long-run average.

Based on the above, a house price decline of 10% by 2009 could be sufficient to restore valuations to a sustainable level. This is consistent with modelling work by academics John Muellbauer and Anthony Murphy, reported in Thursday’s Financial Times.

A larger fall is certainly possible – based on the rental yield, prices have typically undershot their sustainable level by 10-20% during serious economic downturns. However, any such decline would create another attractive entry point for longer-term investors.