As I expected, the RBA’s decision to – rightly or wrongly – slash interest rates to historically very low levels (the discounted rate at which most people borrow is around 6%) has had a near-immediate impact on Australian house prices.
RP Data-Rismark’s hedonic index, which reprices a portfolio of 5 million homes using all historical sales and the incoming flow of approximately 1,500 new sales RP Data collects each day, has recorded a 1.8% increase in Australian capital city dwelling values since the end of May (refer to the blue line the chart below) to July 16. Dwelling values in Melbourne (black line) and Sydney (red line) have risen more sharply, with capital gains of 2.8% and 2.1%, respectively.
Recall that June and July are seasonally very weak months given the tapering in activity over winter. RP Data-Rismark’s daily index, which is published by both the ASX and the RBA (in its monthly chart pack), is produced in raw, non-seasonally adjusted terms since it is used as an investment benchmark. (Tradeabe financial market data are not normally seasonally adjusted.) Yet seasonally adjusting RP Data-Rismark’s index is an instructive analytical exercise, and doing so would show even stronger price appreciation than that which is displayed in the chart above.
Over the period December 31, 2011 through July 16, 2012, inclusive, Australian dwelling values are now more or less unchanged (off -0.4%). In Sydney, dwelling values have actually risen 1.7% over the first six months of the year. And after a steep drop during April and May, home values in Melbourne have recovered quickly to be down 2.4% in 2012. This compares much more favourably with the Melbourne housing market’s -5.6% nadir reached on June 10.
Despite the wealth of empirical evidence supplied by RP Data indicating that dwelling prices have started creeping up again in Sydney and Melbourne, SQM’s Louis Christopher has assuredly argued otherwise. (While SQM competes against RP Data, it does not actually publish a price index of its own.)
Following the release of RP Data-Rismark’s June data, which is being directionally reaffirmed in July, Christopher issued a media release that contained several incorrect statements, which were addressed in a response from RP Data’s Tim Lawless.
One of Christopher’s most surprising arguments, however, was the declaration that, “I do not believe for a moment that house prices are now rising in Sydney or Melbourne as RP Data-Rismark has claimed.” Christopher has been backed up by Sunrise television host David Koch.
Setting aside RP Data-Rismark’s findings, the REIV has separately reported that median house prices in both Victoria and Melbourne rose by about 3% in the June quarter. Averaged over the last six months, the REIV concludes that Melbourne house prices have not fallen at all.
As I explained here, the RBA prudently leverages off multiple house price data providers since it can never be absolutely certain that the quality of any one company’s information will always remain the same. Having said that, the RBA has communicated a clear preference for the “hedonic” and “stratified median” proxies published by RP Data-Rismark and the Fairfax-owned Australian Property Monitors (APM).
Australia is somewhat unique among OECD countries in collecting 100% of all residential sales transactions through valuer-generals and land titles offices and then making this data available to private information vendors. In contrast to the US and UK, where the most widely used indices have large sample selection biases (because they only draw on a sub-section of sales), APM and RP Data-Rismark’s benchmarks ultimately include all sales.
To the extent Christopher genuinely believes that house prices are falling in Sydney and Melbourne, he will presumably defer to APM’s numbers, which will be published late next week.
Christopher publicly says he “created” APM’s stratified median index, which is a point disputed by Fairfax’s APM. In its media releases, APM notes – correctly, I have been informed – that the stratified median index derives from a methodology developed by two RBA economists, Tony Richards and Nalini Prasad (more detail on the dispute between RP Data/APM and Christopher can be found here).
The APM results should be one telling litmus test of RP Data-Rismark’s analysis. While it harnesses similar underlying information, APM employs a very different technical method. I will be personally focusing on the price action APM evidence in Sydney and Melbourne, and the national market more broadly.
The ABS house price index data are much less useful since they completely exclude all forms of “attached” housing – i.e., apartments, townhouses, and terraces – which make up around one-quarter of the housing stock in the major cities. This is why this data is largely ignored by the RBA and most market economists.
Comments