One Tweet that I received told me something that didn’t seem right from my own data: that CPI-deflated US house prices were now within reach of their long-term average (i.e., 1890 till now). From my figures—which I had cobbled together from Robert Shiller’s first edition “Irrational Exuberance” data supplemented by the S&P Case-Shiller Index, adjusted for inflation—prices had fallen a lot, but still had a long way to go.
So I went back to Robert Shiller’s home page, checked his updated data, and sure enough—real US house prices were now “only” 15% over their long term average.
Though this certainly doesn’t mean that house price falls in the US are almost over—having overshot by so much, there is every prospect of them going well below the long term average before they stop falling, especially given the unprecedented mortgage debt from which households are still deleveraging. But it means the day that real US house prices stop falling is closer than I thought before Twitter alerted me to this problem with my data.
The chart below shows the difference between my old series and the new one—which is taken without modification from Robert Shiller’s data page.
Figure 1: Shiller USA Real House Price Data
There revision has intriguing consequences for the Australian market. My Australian data comes from cobbling together Nigel Stapledon’s PhD research into long term real house prices from the late 1800s to the ABS data which began in 1986—and it too is hardly data of the standard that the physical sciences expect. But that said, the Australian house price bubble appears even more extreme: 3 times the long term average, versus “only” twice the long term average when Greenspan told Congress that there was no bubble in US house prices.
And of course, the dominant view in Australia is that there is no house price bubble here. Yeah, right.
Figure 2: Long term real house prices
Even when you consider just the very brief time series that the ABS has assembled, Australia’s bubble dwarfs the American one. Real house prices peaked at 2.6 times the level of 1986, whereas the US market peaked at 1.7 times the 1986 level.
Figure 3: Real House Price Indices from 1986-Now
Though US house prices have now returned to their 1986 level, mortgage debt has not. All the “froth” pumped into the market by 30 years of rising mortgage debt has now gone, but the level of mortgage debt is far higher.
Figure 4: Mortgage debt levels
This is why house prices could fall a lot further in the USA—even though they’re now a relative hair’s breadth from the long term average. As I’ve argued regularly here, the real driver of change in asset prices is the acceleration of debt. I won’t go into the details here—I’m rushing to get this post up before I am on a panel today at the Australian Property Council’s annual gabfest—but I’ll let the next two charts speak for themselves.
Figure 5: Correlation of mortgage acceleration and house price changes, USA
Figure 6: Correlation of mortgage acceleration and house price changes, Australia
Finally, the chart I published some weeks back comparing house price falls from the peaks in the USA, Japan and Australia used the old data, and made Australia’s rate of decline from its June 2010 peak appear comparable to the slow pace of Japanese decline. This revised data indicates that our rate of price fall is more comparable to the US pattern.
Figure 7: House price falls from peak–the early years
How far could Australian house prices go down? Given what has happened to US and Japanese prices, the simplest answer is “a lot more than most of the people in the audience here seem to expect”.
Figure 8: House price falls from peak–the middle years
Figure 9: House price falls from peak–the long term