Updating real house prices
The Nationwide house price index through June shows a decline of about 7% from the peak last autumn. This is being reported as a significant correction but viewed from the perspective of the long history of house prices relative to general inflation, it is only another small step in the market’s four-year defiance of gravity since deflated house prices first peaked relative to their long term trend. Without the oxygen of credit, prices have held up remarkably well but market psychology has changed far more than prices yet indicate. The bear market is about to begin in earnest.
We update in this post our series for the trend of real house prices since 1957, based on the Nationwide data for the ‘average house’ and the Retail Price Index. Earlier posts (search: “real house prices”) explained the significance of the data series as a proxy for fundamental value and as a predictor of future holding-period returns. Real returns matter to occupiers if they are substituting for their own financial asset holdings and they matter if they are borrowing to finance property purchases. They can either rent property or rent money. The relative cost of each will determine whether they build or destroy wealth in real terms. High real property prices in a world in which capital is getting expensive spell real wealth destruction.
The message in real house prices Deflating the Nationwide Index by general inflation (RPI) provides a measure of the real growth in capital values, equivalent to the ‘relative’ inflation in land values. It is more realistic than looking at nominal price indices which mainly reflect the exponential growth of the general price level in the economy. The way prices are normally reported, as an annual rate of change in the nominal index level, tells us far less than real price levels. The chart below shows the deflated series.
This series still suggests exponential growth. However, if we fit a trend to the data we can then express the index as a ratio of the trend. This is shown below. Modifying the data in this way is the same as expressing the actual index in logarithmic terms, so that all movements are proportional.
We have argued in previous items that the trend reflects the sustainable economic growth in capital values, as a function of land availability, relative building costs and income affordability. In the UK, land availability and building costs appear to have less impact on the trend and it is real incomes growth that has mainly dictated the trend. We see the ratio is a measure of value or sustainable future growth, in exactly the same way as long histories of achieved real returns from financial assets are used in our long-term asset model as a predictor of future real returns. Deviations above trend point to falling real prices and vice versa.
Exceptional overvaluation The level of overvaluation in this cycle appears extreme but not exceptionally so. There are two reasons why this is misleading and why the overvaluation is probably unprecedented.
First, if ‘hindsight’ were omitted from the data, and in each cycle only data up to that point were known, the recent ratio relative to the highsight-free trend is more extreme than in the bull-market peaks of 1972 and 1988.
Second, the fitted trend today is 2.8% in real terms but up to six years ago, when we first wrote up this approach for the FT, the trend was about 2%. This may simply be telling us there is not enough history (or enough cycles within the history) to remove the effects of the most recent up cycle. If, after the next down cycle, the trend drops back to 2% or so, then the recent deviation will show up as much more extreme than in previous cycles.
It is possible the sustainable trend has genuinely increased from 2%. How might this arise? Possible reasons are – if land availability has reduced (the reverse appears to hold) – family formation is changing and increasing demand relative to the population (which does hold); or – building cost inflation has risen risen relative to general inflation and will not fall back again (it has not, although likewise there have been no technological breakthroughs to lower relative building costs).
The Nationwide Index, incidentally, corrects for changes in the quality of the housing stock over the years, which it rightly treats as a further investment in housing rather than as a gain. Failure to make thsi distinction is one reason the trend real growth is so much lower than most home owners think, although as important is the fact that most people cannot hold in their heads the cumulative change in the general price level in the economy over a long period of home ownership.
Our view is that it is sensible to assume that the sustainable trend will fall back after the next bear market to close to 2% again, in which case the implied future returns from housing are even lower than if we treat the current over-valuation as about 30%.
Market timing In common with our measures of value and implied future return for financial assets, this approach to housing returns does not tell us about the near term. Timing requires different analytical or judgement tools.
Readers of this blog will know we regard market timing skills as rather implausible. This is borne out by the fact that, in this housing cycle, the extreme deviation has persisted for much longer than in past cycles, having first reached its peak in 2004. Though not predictable, it reflects monetary policy failures whose roots lie in a mistaken belief that inflation and the business cycle had both moderated and that rampant credit growth and rapid house price inflation did not matter. Ironically, politicians seem deliberately to have erred on the side of avoidance of a correction even though it would have been infinitely more useful socially than the bubble we have instead experienced. Lower prices would have helped affordability and prevented poor decisions about borrowing and consumption, including equity withdrawal by both younger consumers and elder retirees. It is now too late to correct these without widespread harm to living standards.
In spite of our scepticism about market timing, it is now very clear that the market cycle has turned down and that, without a miracle cure for bankers’ difficulties funding their loan book, financing conditions are going to remain very restrictive. But what also makes bear markets is investor psychology. This has clearly started to shift as the most intransigent housing bulls wake up and smell the coffee. It helps that the impact of the credit crunch is being seen across the retail spectrum and not just in the housing market. But it is still not being driven by fear of income and job losses. That may follow soon.
Market illiquidity In every cycle there is some feature that is different from the past. This is also likely to shape psychology. We are beginning to see that this might be illiquidity.
Liquidity, evidenced by transaction volumes and frictional costs of moving, can dry up as sentiment first shifts from bullish to bearish. Sellers look back to the highs and are at first reluctant to discount selling prices. Eventually, discounted prices establish a new, lower level and create more widespread expectations of a bear market. Sellers anticipate further ratcheting down and so are willing to discount to achieve a sale. At this point supply increases and liquidity improves. One year after the collapese of Northern Rock, this is where we might expect to be now. In fact, liquidity is now being constrained more by buyers than sellers, as access to credit has moved almost overnight from conditions of extreme laxity to lending criteria that have not been seen for several decades.
Illiquidity increases the frictional costs of moving home, by widening spreads between selling and buying prices. Frictional costs had already been mistakenly increased by levying stamp duty, as yet another supposedly harm-free stealth tax. Non-interest costs have also been increased by lenders in an attempt to offset sharply higher funding costs. These affect flexibility to move homes but also refinancing flexibility, notably as fixed rate periods end and borrowers find they effectively no longer have an option to change lender.
Illiquidity could be as harmful as weaker prices, by preventing actions that households need to make either as a function of general economic developments (needing to release excess equity or reduce excess debt) or family-specific developments (having children or changing job location).
As in other asset markets, illiquidity is likely to increase price declines, as long as some transactions have to occur. In housing, they do.
Downside risks Previous bear markets in real house prices, peak to trough, were 25%, 17%and 47%. Because of high general inflation, the declines in nominal prices were much lower. This time round, the general inflation rate is unlikely to be so high and the overvaluation appears even more extreme. Even without forming a particular view of the economy, such as personal incomes and unemployment, it is sensible to plan on downside risk being greater, in real terms, than in the three previous cycles.
How long it persist might be as important as how low prices go. Previous posts on the lessons of Japan point out that a bear market in relative property prices can last up to 15 years.
The key is not just duration and degree but the opportunity cost in terms of real interest rates, or the cost of renting other people’s money. There is a growing awareness that we have moved from a world economy dominated by a small number of high-income countries to a more multi-polar economy with far greater competition for savings capital. This is why the gap between the achieved capital returns on housing and borrowing costs may expand in an unprecedented way, causing massive (even if gradual) wealth destruction.
We are conscious this is not a very sophisticated argument but sometimes it is smart to be simple. Correction: most times.