Few measures of the housing sector are more valuable than loan approvals and dwelling turnover. Both measures -- operating as a proxy for housing demand -- provide important insight into the health and potential direction of the property market.
Property upswings are almost always associated with a strong rise in lending and property sales. The relationship doesn’t always hold as cleanly as analysts would like -- shifts in the composition of buyers and lenders can temporarily affect the relationship between prices and sales. But the occasional false signal is a small price to pay to identify shifts in Australia’s $5 trillion housing market.
Loan approvals and dwelling turnover track each other closely but not without some significant complications. Dwelling turnover data is not particularly timely, owing to the considerable lag between when a sale is made and when it is officially reported by each state’s valuer-general.
Useful estimates of dwelling turnover are typically available about three months after the reference period but it often takes more than six months before all sales are officially reported. Furthermore, reporting differs significantly by state; for example, property sales in South Australia are collected and reported more efficiently than in Victoria.
As such, the most timely measure of housing demand is the Australian Bureau of Statistics’ monthly lending finance data. Unfortunately, mortgage approvals have their own issues.
First, not every property buyer takes out a mortgage. According to the HILDA survey, between 20 and 30 per cent of all property sales are conducted without taking out a loan. Obviously those figures differ considerably with age; for example, only 20 per cent of purchases by households where the main breadwinner is over 55 years old feature a mortgage.
Second, there is considerable double counting in estimates of loan approvals. It is not unusual for prospective buyers to receive loan approvals from a range of banks before making a decision. In theory, these duplicates should be removed from the data but in reality this happens in a very ad hoc manner.
Third, just because you receive an approval doesn’t mean than you actually receive a loan. Cancellations can and do occur, creating some bias to the data.
Fourth, there is no measure for the number of investor loan approvals. The ABS only reports the value of investor loans but it isn’t too difficult to come up with a reasonable estimate.
Fifth, it does not capture foreign purchases. This category has been rising in recent years -- although legally foreigners are not allowed to buy existing property -- but on a whole it remains a fairly small share of total activity.
On a whole double counting and cancellations create some upward bias to the lending figures, which is partly offset by the fact that about 20 to 30 per cent of property sales do not involve a loan. Despite the competing biases, loan approvals and dwelling turnover track each other pretty well, as the graph below shows.
On a trend basis, dwelling turnover peaked late last year and has eased somewhat through to March. Early estimates of turnover in April and May suggest that demand may have soften a little more.
The number of loan approvals has also eased in recent months, although the value of loan approvals continues to rise. Investors have driven most of the recent growth and continue to underpin the property upswing.
Both indicators suggest that the pace of house price growth has peaked, with the current upswing relatively minor compared with the 2009 boom. Consistent with this, annual house price growth eased to 10.1 per cent (from 10.9 per cent in the March quarter). Price growth moderated in each mainland capital besides Brisbane.
The data suggests that most of the risks to housing lie on the downside. Lending activity has been elevated and historically has struggled to maintain this type of momentum for long.
But the limitations of loan approvals as a measure of demand may provide some upside risk for prices -- could greater foreign demand keep the market ticking along? Are prospective buyers shopping around to get the best deal? What about self-managed superannuation funds?
Furthermore, the recent property boom has been centred in Sydney and Melbourne. It is possible that stronger growth in the other three cities could offset the expected decline from Australia’s two biggest cities – although for now the Perth and Adelaide markets remain weak.
I suspect that none of those factors will stop prices declining significantly over the next couple of years, with Brisbane poised to be the best performing city. What remains uncertain is whether that correction will be similar to recent downturns -- nominal prices down 5 per cent; real prices down 10 per cent -- or part of a more significant correction as Australia adjusts to life after the terms-of-trade boom.
Loan approvals and dwelling turnover provide a great deal of insight in the cyclical triggers for the housing market. At the moment they suggest that the market is close to its peak and may start to ease before the end of the year.
However, we shouldn’t place too much faith in these measures because they don’t, for example, provide any insight into the longer-run drivers of the property market. For that we need to assess the outlook for factors such as wage and income growth, demographics, credit outstanding and risk preference.