How stretched is the UK housing market? Given how long UK house prices, especially those in London, have defied both gravity and conventional models of valuation, a confident answer to that question is hardly possible. A look at a couple of the conventional “affordability” measures suggests, however, that – in MPC speak – “the balance of risk to house prices is skewed to the downside”. In English, house prices are more likely to fall sharply than to rise steeply.
Exhibit 1 is a measure of average UK house prices to average earnings. The specific measure is Nationwide’s ‘First time buyer gross house price to (annual) earnings ratio’. As shown in Chart 1, the current value of this ratio, 5.3, is the highest since 1983 (in fact since records began), easily exceeding the 3.9 reached during the previous housing bubble in Q2 1989. The average value over the period shown in the Chart 1 is 3.2. If we fit a simple linear trend over the period, the most recent value of the (rising) trend line would be just under 4.0.
There is no logical reason for the house price to earnings ratio to revert to either its historical average value or to its trend value. Even if we believed that a reversal to trend, say, were inevitable, this would not tell us whether this would happen through a decline in house prices, through a rise in earnings or possibly even through a further increase in house prices swamped by an explosion in earnings growth. Earnings growth has been much more stable, however, than house price inflation. So let’s assume earnings will continue to grow at, say, 4.5 percent per annum. To restore the house-price to earnings ratio to its historical average of 3.2 over a two-year period would require almost a almost 19 percent annual decline in house prices (for each of these two years). If the restoration of the historical average ratio were to take three years, the annual rate of decline of house prices would be just under 12 percent.
If the house price to earnings ratio reverted not to its historical average but to its 2007 Q2 trend value of 4.0, the annual rate of decline of house prices would be around nine percent if the 4.0 value for the ratio were to be reached in two years, and about five percent if it were to take three years.
Of course these are just ‘what if’ calculations, but these and similar, more detailed numerical simulations do cause me as someone who is ‘long housing’ (that is, someone who owns a house whose value is greater than the present discounted value of the future housing services I plan to consume) to sit up and take notice.
A similar worrying story (for home owners) is told by Exhibit 2, shown in Chart 2 below. This plots Nationwide’s First Time Buyer Affordability Index, mortgage payments as a percentage of take home pay since 1983. This index peaked in 1989 Q2 at 147.6 percent. Its current value, 133.4 percent is still some way below the all-time high, but well above the average for the period of 85.2 percent. While the Official Policy rate set by the MPC is now coming down, and could end up quite a bit lower than its current 5.5% value (say around 4.5% a year from now), the cost of mortgage borrowing is unlikely to come down much, if at all, even for those who meet the much stricter eligibility requirements imposed in the wake of the 2007 financial crises. There won’t be much aggressive mortgage borrowing or lending any time soon to prop up the UK housing market.
All this affordability analysis is, of course, rather woolly stuff that an economist approaching housing valuation using a conventional finance approach would not deign to touch. A finance perspective sees the value of a house as the present discounted value of current and future rental earnings, actual or imputed, derived from the ownership of the house. With a bit of hand waving, this produces the following expression for the ratio of the current house price to the current rentals (actual or imputed) from housing:
where p is the current house price – rental ratio, r is the long-run risk-free interest rate, ? is the long-run average risk premium on housing equity and g is the long-run expected growth rate of housing rentals. Finally, b is the value of the housing bubble (if there is one). The first term on the right-hand-side is therefore the ‘fundamental value’ of the price-to-rental ratio.
Market measures of the long-term risk-free real rate of interest for the UK continue to be ludicrously low, partly because of continuing anomalies in the demand for long-dated index-linked gilts by pension funds and partly because of the inexplicable failure of the UK authorities (and other public agencies) to issue more of the stuff. There can be little doubt, however, that the real risk-free rate is rising globally and also in the UK. The decline in the real long-term risk-free rate of the early years of this century no doubt accounted for part of the increase in the fundamental value of UK property. That process is now in the process of being reversed.
It is hard to tell a convincing empirical story about the fundamental determinants of the housing equity risk premium, and I will not try to do so here. The future expected growth of actual or imputed rental income from housing therefore must carry quite a bit of the weight in any rationalisation in terms of fundamentals of the past and present strength of UK house prices. A limited supply of new residential housing (due in part to planning and zoning restrictions) other than in parts of the buy-to-let market and, in comparison with much of continental Europe, relatively housing-friendly demographics, were sources of housing price strength. With suitable land in highly inelastic supply and real incomes rising, rents could well grow faster than earnings almost indefinitely. The fact that London is the favourite pied-a-terre for every nouveau riche and ancien riche alike, has influenced house prices well beyond the rarified quarters where the owners of the mega money themselves are squatting. I don’t compete in the housing market with Roman Abramovitz, but I may well compete with his chauffeur, or with the child minder of his chauffeur.
The fact that prices in part of the UK housing market are strongly influenced by global fads, fashions and other factors is good news for UK homeowners when the UK economy is about to slow down significantly, as it is now. It is, however, also a potential source of vulnerability. Fads and fashions change. Tax-averse expatriates may be footloose and may consider a move to Paris, New York or Dubai when their non-domiciled tax status in the UK is threatened. I also believe that even the mega-rich will at some point discover that London has no transport infrastructure, public or private, and may as a result consider relocation. One dirty bomb, even a small one, would certainly end the party once and for all. I am not making predictions here – just listing risk factors.
So yes, I consider a significant decline in UK house prices, say by thirty percent or so over two or three years, to be likely but not inevitable.
Would it matter, and if so, would it be a good thing or a bad thing? On balance a good thing, I would say, although I myself would stand to lose quite a bit. A decline in house prices does not, on average, make UK residents worse off. It merely redistributes wealth from those long in housing to those short in housing: the representative UK household will lose as homeowners, but will gain the same amount in present discounted value terms, either as renters or through the opportunity cost of the imputed cost of renting saved by owner-occupiers. Typically the middle-aged and the old who have not yet traded down will lose. The young and all those expecting to trade up will gain. Since housing prices in the UK have become ridiculous (not just in London either) and an increasingly large contingent of first-time would be buyers is priced out of the housing market at current prices, a significant decline in house prices would, on balance be welcome. Just imagine: a London where essential workers (police, fire fighters, teachers, nurses, bus drivers etc.) can actually afford to live! I wonder what such a city would look like. If the market can achieve this kind of socially desirable redistribution, who knows, governments may discover ways to enhance efficiency.
There are two downsides to this scenario, in addition to the losses suffered by those long housing. The first is that if the decline in house prices reflects the bursting of a bubble (a decline in b in the formula) rather than a decline in the fundamental value of residential property, there will be a net negative wealth effect, because there will be no consumers of housing services to benefit when the value of the housing stock declines. This negative wealth effect could be macroeconomically significant.
Second, even if there is no net wealth effect from a decline in house prices, there will be a liquidity or collateral effect. Residential property is mortgageable, unlike human capital. It acts as security for consumption loans (through mortgage equity withdrawal) that would not be forthcoming if they had to be provided on an unsecured basis. When this source of collateral shrinks, consumer borrowing will decline and consumer spending is likely to decline with it.
These detrimental effects from a decline in the price of residential property are, however, likely to be transient – cyclical at most. The social benefits from a significant decline in UK house prices are bound to be significantly larger than the social costs associated with any short-term credit squeeze on consumers it may cause or aggravate.