prices double every eight years, you need to do little more than find a property in your price bracket. The location is not crucial or even important because, once a property is secured, you only need to hold and hope. But surely, if we are going to base future housing performance on the past performance of the market, we should first check to see whether our current economic, social and financial conditions are roughly the same as those prevailing during the period we are using for comparison.
Figure 1.1 : house prices doubled in value every eight years from 1967 to 1990
Source: Australian National Library's online Trove facility and Mitchell Library archives.
When we do this, we find that the years from 1970 to 1990 were not typical at all, and certainly nothing like what we are experiencing or likely to experience in the foreseeable future. The 1970s and 1980s witnessed the highest inflation levels in our history; many of those years had double-digit inflation, which was responsible for much of the house price-growth. There was little real price-growth. Those who analyse the past to predict the future ignore this fact completely. Maybe they don't really understand the long-term performance of the housing market at all, or just maybe they promote this theory because it gets them off the hook. Imagine you buy a property and the price starts to fall, and keeps falling for several years. What happens if you complain to whoever advised you? They'll tell you to wait, and assure you that growth will come because it always has in the past. If prices keep falling, they may hope that you'll forget who it was that gave you such bad advice and they'll be off the hook completely.
This focus on past performance is also pushed by project marketers and developers, who use high past price-growth as an indicator of expected growth. ‘Get in quick,’ they'll tell you, ‘or you'll miss out. These properties are selling like hot cakes.’ But does past performance guarantee future growth? Some of the best-performing suburbs and towns from 2000 to 2010, where house and unit prices regularly rose by well over 10 per cent per annum, such as the Gold Coast, Mackay, Gladstone, Moranbah and Bowen, suffered massive price falls in the following years. Investors who bought in those towns at the peak of the boom in 2010 then watched in dismay as rental vacancies shot up and prices plummeted by over 50 per cent. Relying on past performance to predict the future is like trying to drive a car by looking through the rear-view mirror. It's a good method to see where you've been, but completely useless at showing you which direction the road ahead is taking.
What we've learned from historical data
The only way that past performance is going to tell us anything at all about the behaviour of the housing market is if we go back not just forty or so years, but far enough to see how it performed through economic booms, recessions and depressions, periods of high and low inflation, war and peace, easy housing finance and no finance, high population growth and low growth. In short, we need to go back in time as far as the data allows us. Luckily some academics have done just that and we can share in what they discovered. Here are their studies:
► Stapledon, Nigel, Long Term Housing Prices in Australia and some Economic Perspectives. Sydney: University of NSW, 2008.
► Eichholtz, Piet, M. A., A Long Run House Price Index: The Herengracht Index 1628–2008. Maastricht: Real Estate Economics, 2010.
► Lindeman, John, Mastering the Australian Housing Market. Melbourne: Wrightbooks, 2011.
► Conefrey, Thomas, and Karl Whelan, Demand and Prices in the US Housing Market. Dublin: Central Bank of Ireland, 2012.
The Herengracht study analysed nearly four hundred years of house price movements in Amsterdam, while Stapledon's study at the University of NSW researched over one hundred and twenty years of house price changes in Sydney. My own study, published in my previous book, Mastering the Australian Housing Market, covered Australian capital city house price movements from 1901 to 2010, while Thomas Conefrey and Karl Whelan analysed the performance of the US housing market from 1968 to 2012. All of these studies came to three very similar and quite amazing conclusions about how the housing market performs over long periods of time. The following pages summarise these findings.
Figure 1.2 shows the performance of the Australian capital city housing market since 1903, and it provides a far more sombre picture than that promoted by the buy and hold advocates.
Figure 1.2 : house price movements every eight years from 1903 to 2014
Source: Australian National Library's online Trove facility; Mitchell Library archives.
Since 1903, there have only been four times when house prices doubled in price over eight years. Far more significant is the fact that every other eight-year period has failed to meet this performance benchmark. Not only have there been lengthy periods with little price-growth, house prices actually fell during the 1930s and failed to rise during the 1950s. The price-growth achieved by houses in Australian capital cities every eight years has been about 55 per cent, which means that, on average, it takes 13 years for house prices to double, not eight or even ten.
Unfortunately, even that information is of no use when it comes to predicting the future, because the rate of house price-growth has been highly irregular. It all goes to prove that:
► past performance does not predict future performance
► there is no housing market cycle or property clock
► housing prices have not doubled in price every eight or even ten years
► these common assumptions about the housing market's performance are not based on facts, but on assumptions that can be quite misleading.
Luckily for investors, the studies have also identified the cause of this apparent unpredictability.
Much of the irregular price-growth of housing is caused by the underlying inflation rate, and it was high inflation that led to the well-touted doubling of house prices every eight years in the 1970s and 1980s. Figure 1.3 (overleaf) demonstrates this correlation, with house prices closely copying the inflation rate, but usually performing slightly better.
Figure 1.3 : house price movements mirror the underlying inflation rate
Source: Australian National Library's online Trove facility; Mitchell Library archives; CPI and IRPI data adapted from Data on Request, Australian Bureau of Statistics.
In fact, the only years when house price-growth significantly exceeded the rate of inflation was during the postwar baby boom years of 1947–50, during the years 1970–73 when the baby boomers purchased their own houses, and then again during 1998–2002 when the children of the baby boomers purchased their homes.
The close correlation between house prices and inflation means that the behaviour of the housing market is actually quite predictable, but only when measured at a national level and only over long periods of time. House price-growth in major cities has consistently averaged about 2 per cent per annum above the underlying inflation rate since 1901, but since the end of the Second World War in 1945 this rate has risen to about 3 per cent per year. The cause of the slightly higher growth rate is due to higher demand for housing in capital cities. Our overseas migration rate has been much higher since the war, and most of these new residents prefer to live in cities such as Sydney and Melbourne, resulting in a higher demand for housing than in regional and rural markets.
High demand for city living is unlikely to change anytime in the foreseeable future, giving us a simple rule of thumb to predict price-growth. Over long periods of time, the rate of house price-growth is likely to average about 3 per cent in capital cities plus the underlying rate of inflation. This means that when the annual rate of inflation averages about 2 per cent, a typical investment property is likely to