The Australian property sector is into the ‘second quarter’ of what is often a seven-to-eight-year housing cycle, according to Stockland chief executive Mark Steinert. Unfortunately, the average property cycle has been a lot closer to four years over the past two decades, which means that the ‘second quarter’ is a lot closer to the end.
Residential construction has been a bright spot for the Australian economy over the past couple of years and Stockland believe that the good times are just getting started.
“Volumes and activity will stay high for some time to come,” Steinert told Fairfax. “Because all cities are undersupplied.”
From a structural perspective, Stockland is perhaps better positioned than some of its rivals, with 42 per cent of their customers first home buyers: a segment of the market that is likely to improve somewhat as demand from investors begins to ease.
However, there is ample evidence that new construction will begin to slow next year and that slower population growth could see the supply gap closed a lot quicker than commonly believed. Furthermore, the construction market clearly moves in cycles closer to four years in length rather than the seven-to-eight years indicated by Steinert.
There have been five downturns in residential construction since 1994 as measured by the peak in building approvals. The length of these episodes varies but it averages out to a downturn every four years.
More recently building approvals appear to have peaked in the March quarter this year. The trend since then has turned negative and based on previous cycles we should expect approvals to fall rapidly from this point.
The cycle for completions tends to be less volatile -- partly because approvals don’t always translate into activity -- but again the average cycle is much shorter than the seven-to-eight years indicated by Steinert.
The longest upswing for completions occurred between June 2001 and September 2005, although there were a couple of mini-cycles during that period. The current upswing is three years old and based on the historical relationship between approvals and completions -- and assuming that approvals fall further from this point -- construction will likely peak in either March or June next year.
The cycle for building activity tends to be similar to that of house prices. There have been three property downturns since 2003 of varied size and length, which indicates that prices aren’t as sticky as is often believed.
With APRA actively trying to curb investor lending -- and wage growth remaining at a level normally associated with a recession -- price growth should begin to moderate in Sydney and Melbourne. Prices in other cities remain weak and there is considerable risk in Perth as mining investment collapses and export income falls short of expectations.
Although there has been a clear undersupply of housing over the past couple of decades, that hasn’t stopped prices from falling during periods when the demand and supply dynamic shifted. A mining and income boom and favourable demographics didn’t stop house prices from falling in 2008 or 2011.
It’s also worth noting that the population-to-residential construction ratio has fallen towards its lowest level in around a decade. At this level, the housing gap is beginning to close -- and at a relatively rapid pace.
Another way to look at this is to trace what tends to happen when building approvals pass their peak. In the graph below I have identified the five peaks in building approvals since 1994 and traced the pace of deterioration for building approvals and what that means for commencements and completions.
Building approvals and commencements tend to follow a very similar trend. However, the trend for completions can often be quite different due to a combination of the lag with which it takes to complete a property and the fact that some approvals and commencements are simply not completed.
In three of the five scenarios, completions peaked three quarters after the peak in building approvals. On two occasions, residential construction remained relatively upbeat even as the rate of growth stalled. In three episodes, residential construction fell significantly.
The recent construction boom has been one of the largest on record and suggests that the peak will be followed by a relatively large decline. Complicating matters somewhat is that the market has been dominated by construction of higher-density inner city apartments which tend to take longer to complete. This could extend the cycle by maybe a quarter.
Either way the idea of a seven-to-eight year property cycle isn’t supported by the facts. Perhaps Stockland’s internals disagree -- perhaps their market segment is less volatile than the broader market -- or perhaps they are simply optimistic because that’s the nature of property groups.
But based on the market over the past 20 years -- and with knowledge that there is nothing particularly unusual about this episode -- it appears as though residential construction should peak in either March or June next year.
The reality is that the ‘second quarter’ is actually a lot closer to the final quarter. Activity might remain at a somewhat elevated level through to the end of 2016 but will most likely subtract from economic growth. In turn, this should weigh on the profitability of property developers and firms exposed to the property sector.