Housing's Low Point

The graph below shows how Australian dwelling approvals have moved in the period between 1983 and 2006. What we see over the 23 years is that there are four tops and five lows in the housing cycle. What that tells you is that the average length of the Australian housing cycle is about five years. The last low in dwelling approvals was in the middle of 2001. That tells you that the next low in the housing cycle should be on average some time around now, in the middle of calendar 2006.

(The series for Australian dwelling approvals exhibits nine cyclical turning points including peaks and troughs over a period of 272 months from July 1983 to February 2006. The average length of the cycle can then be found by multiplying that 272 by two then dividing by nine. The result, 60.44 months or five years.)

The level of the lows of the previous cycle was 8953 dwellings approved per month. The peak was about 18 months later at 19,590 approvals per month. There are now about 11,450 dwellings approved per month. That tells you that we are below average but not near the levels you would expect in terms of the absolute lows.

The other thing we note when we look at the levels of the lows is that the market usually hangs around the bottom for around about a year or so. The American satirist PJ O’Rourke has described the US economy as being in two stages, “okey dokey” or “not so hot so”. What you see about the dwelling market is it is either flat out or dead flat and it seems to flick from one to the other pretty rapidly. We are just beginning to come in to a bottoming period.

What drives dwelling approvals in the long term is full-time employment, because it is only people with full-time jobs who can pay off mortgages. In Figure 2 we see the Australian full-time employment over the period from 1983 to 2006. What we see is over those 23 years the number of full-time jobs in Australia has risen from about 5.2 million back in the mid 1980s to about 7.2 million now. Full-time jobs growth has been strong. The trend has been going up.

We now look at the level of dwellings approved compared to the number of full-time jobs. When we do that we can demonstrate the housing cycle. What we find is that the market tends to peak at about 2.8 dwellings approved per 1,000 people in full-time jobs and it seems to bottom out about 1.4. The long term average ratio is 2.1. The third graph shows that the level of dwellings approved and full-time jobs showed a percent difference from the long-term average.

There are a couple of things we can say about this cycle. First, as we have said before, the average period seems to be five years. The slump in 2001 was a bit deeper than in previous cycles. The recovery rose briefly to the previous peaks. This cycle didn’t appear to provide the level of high production or over-building that was typical in the mid-1990s or the late 1980s. That may explain why the response was in terms of price, not quantity. The fact that the supply response was less probably explains why the price response was greater. Right now we are down around about 1.7 dwelling approvals per 1000 people in the workforce and that compares to the average of 2.1. That means we are 19.6% below average. The previous cycle low in 2000 was 37.1% below average.

The other thing that drives the housing cycle (as everybody knows) is interest rates. We find that real short-term interest rates seem to drive dwelling approvals by around three quarters. Usually the bottom of the housing cycle occurs three quarters after the final increase in interest rates.

Now we have just had an increase in interest rates, and the negative effect on those dwelling approvals will be around the end of this year in around three quarters.

When we get down to this area at the bottom of the market, which PJ O’Rourke might call “not so hot so”, it takes a little over a year for the market to bottom out. We are just entering this period.

We should have seen the bottom of the housing cycle by the end of calendar 2007. We are now approaching the bottom of the housing cycle. By the end of calendar 2007, we should be lifting off the bottom into the next major housing recovery.