How much is the housing market cooling? HSBC's Paul Bloxham
GUEST OBSERVER
There is significant local debate about how much the Australian housing market is cooling.
Various indicators are showing different trends. A prominent hedonic house price measure is showing stronger growth than the broad collection of housing price indicators. Auction clearance rates have also been high, suggesting price strength.
At the same time, new housing loan approvals are below their 2015 peaks, housing credit growth has slowed and housing turnover has fallen. We see the collection of evidence supporting the idea that the housing market has cooled (and so does the RBA). This is welcome, after an extended period of strong housing price growth in Sydney and Melbourne.
Our central view has the RBA on hold from here. If the housing market proves to be more resilient than we expect, this would further bolster that view.
Australia has a collection of housing price indicators. These indices take different approaches to dealing with the key measurement challenge, which is that there is significant variation in the sorts of houses that are sold month to month. For example, if a lot of large, well-located and expensive houses are sold in one month and smaller, less well-located, cheap houses are sold another, this can have a big effect on the raw median price measures.
In this case, housing price growth would appear to slow, but this would not be a true picture of the change in price of the overall housing stock. Instead, it would be a reflection of the compositional changes in what was being sold month to month.
The four main measures tracked in Australia that seek to control for these compositional effects are a hedonic measure (Corelogic), a stratified median measure (APM), a repeat-sales measure (Residex) and the official ABS measure, which also uses a stratification approach.
Hedonic measures control for compositional shifts in turnover by accounting for things like number of bedrooms, block size and location, amongst other characteristics. The stratified approach groups houses into ten different price level boxes and then averages the medians from each box to produce an overall measure. The repeat-sales method controls for quality by tracking the same houses through time.
At the moment, the various measures are showing y-o-y growth in national housing prices of between 7 percent (Corelogic) and 3 percent (APM) (Charts 1 and 2). On a q-o-q basis, the gap is even larger, with the hedonic measures showing more strength recently. Keep in mind that the main debate is really only about whether the Sydney and Melbourne markets are cooling, as there has been little housing price growth in other cities in recent years (Chart 3).
The gap between the measures has led to some debate about which measure is closest to the truth. A challenge has been that the Corelogic measure, which is showing the strongest growth and has typically been the one most relied upon by the RBA in recent years, has recently had some methodological changes. As a result, the RBA has become less confident about the veracity of this measure, stating in its August official statement that 'strong increases reported by CoreLogic were overstated as a result of methodological changes'.
Digging a little deeper into the issue (to the extent that we can without the unit records), we find it interesting that the CoreLogic raw median, which does not adjust for compositional changes, has also shown much slower housing price growth in Sydney than the adjusted measure (Chart 4). The gap between the hedonic measure and the raw median is the largest it has ever been. The clear implication is that compositional shifts in turnover have been quite large recently.
One plausible explanation is that a greater amount of lower-priced housing has been turned over recently. This fits with the fact that a lot of newly built apartments have been coming to market and apartments tend to be lower-priced than detached houses. Apartment prices have also grown more slowly than detached dwelling prices more recently. Lower-priced detached houses are also the ones that are likely to see the biggest spill-over effect from the apartment supply boost, as they are typically a closer substitute to most apartments than higher-priced houses.
Another timely measure of housing market activity, auction clearance rates, has also been high in recent months (Chart 5). This may also be consistent with the gap in the housing price measures. Auctions are more prevalent for higher-priced homes and account for only 30% of housing sales in Melbourne, 20% in Sydney and far lower rates in the other capitals. Auction volumes have recently been lower than in recent years, suggesting fewer high-priced dwellings are being sold.
More sales of lower-priced dwellings would also be consistent with the fall in housing loan approvals and slowdown in housing credit growth, as access to credit is typically a more important factor for lower- priced housing (Chart 6). The tightening of lending standards since late 2014 has also applied most aggressively to loans to investors, which are more prevalent in the apartment market.
The bottom line is that the weight of evidence suggests that there has been some cooling in the housing market (for more details on why the market it is likely to have cooled, see: 'Downunder Digest: Australian housing is still set to cool', 21 June 2016). Compositional shifts in housing turnover are making it difficult to get a clear read on housing price growth, but, importantly, housing loan approvals and credit growth have slowed. For monetary policy, the slowdown in housing credit growth is particularly important. Housing price gains are far more worrisome when they are fuelled by credit. It is the leverage that matters most.
Our central case is that the RBA is unlikely to cut further, with this call grounded in our view that domestic inflation and wages growth is set to stabilise as the rebalancing act comes to its end (see 'Australia's low inflation: It's part of the rebalancing act', 13 July 2016). If housing price growth proves to be stronger than most of the measures show, or there is a re-acceleration of housing prices, loan approvals or credit growth, we see this as likely to bolster, rather than challenge, our case that the RBA is unlikely to cut further.
PAUL BLOXHAM IS CHIEF ECONOMIST (AUSTRALIA AND NEW ZEALAND) FOR HSBC.
