The Bank is losing a house price game

First published in the Financial Times, UK edition, by Andrew Farlow, 29 November. The writer is the tutor in economics at Oriel College, Oxford University.

The chancellor's argument in his pre-Budget report that the housing market is "sustainable" rests on the claim that it has not experienced a bubble. But housing markets are just as prone to bubbles as other asset markets, if not more so. Rising prices draw buyers in and encourage owners to increase their debts. This pushes prices up further and the cycle continues. Even for those who know of the bubble, their ability to bet against it is limited. Owners are illiquid. Timing exit and re-entry is impossible. New entrants betting on the market's going up swamp those betting on its going down.

Those who put rising prices down to "fundamentals" are wrong. What connection between changes in fundamentals and house prices could generate a 30 per cent rise in one year? House prices have risen greatly even in regions where fundamentals have not improved. This shifts the onus of proof on to macroeconomic factors - but there are none to explain such a rise and many that point in the opposite direction.

We are told that, since the average ratio of debt service to income is low, all is well. It is not. The efficient ratio should be based on the expected interest rates and incomes over the full life of loans, so we would expect the current ratio to be unusually low. And it is the weakest, not the average, debt-holders who drive crashes; the more skewed the distribution of debt, the more fragile the bubble. In the 1970s and 1980s the real burden of debt was reduced by inflation; low interest rates today make it seem easier to service but the real burden is higher. Last, much recent lending, on so-called "affordability measures", is flawed; it ignores uncertainties such as the collapse of the bubble itself. Instead of offsetting the borrowers' lack of sophistication, banks - competing for customers - have added to it.

An interest rate trigger is not necessary for a crash. If interest rates had not peaked in 1990, that bubble would have just got bigger and burst in response to a smaller trigger later or to no trigger at all. Imagine if UK house prices stopped rising. Banks would be less willing to lend and borrowers less willing to borrow; house prices would start to fall; consumers would try to replenish savings; the collapsing bubble would generate negative expectations; consumption, investment and the stock market would fall; unemployment would rise and, if the pound fell, there would be inflationary pressures; interest rates would rise, increasing the burden of debt and putting further downward pressure on house prices.

Mervyn King, now governor-designate of the Bank of England, has inadvertently revealed that the Bank now thinks we face a bubble. This month he said house price growth would slow to "roughly zero" in two years' time but "the probability that the out-turn will be close to the central view is zero". A probability distribution centred on zero but with all the probability in the tails? A perfect description of a bubble! Prices may yet go higher, but may also collapse.

Mr King also said: "The best way to destroy the credibility of the monetary policy committee is to lecture people about house prices." Ever since Alan Greenspan's famous "irrational exuberance" warning, central bankers have reasoned that, to avoid looking stupid, they should not warn people about an asset price bubble in case the bubble gets more extreme before crashing.

The truth is that the Bank has been playing a risky game with the housing market: trying to set interest rates just low enough to stimulate consumer spending but not so low as to set off a house price bubble. It is a strategy that has gone wrong.