On Thursday September 9th 2010, the Commonwealth Bank released a document on the Australian housing market, to support a tour that its senior executives are making to meet overseas investors. The press release for the document was as follows:
Australian Residential Housing
Sydney, 9 September 2010: Senior executives from the Commonwealth Bank of Australia (“the Group”) will soon be travelling overseas to meet with some of the Group’s offshore shareholders and other investors interested in Australia and the Australian banking sector.
In the light of recent commentary from a number of sources on the robustness of the Australian residential housing market, the Group (given its significant exposure to this section of the economy) anticipates that this will be an important issue for many of the investors it is scheduled to meet with.
In anticipation of these discussions, the Group has produced a presentation entitled “Australian residential housing mortgages: CBA mortgage book secure”. A copy of this document has been lodged with the ASX today. (Commonwealth Bank)
The document the press release refers to has a table (on page 4) comparing house price to income ratios in Australia with those from a number of coastal cities overseas–with the argument being that Australia’s relatively high house price to income ratios are a by-product of the fact that Australia’s major cities are all located on the coast. By the looks of this table, Australia’s house prices are comparable to those elsewhere–Sydney’s 6.2 ratio, for instance is below the 7 times ratio of its sister city in the USA, San Francisco.
So there’s nothing for overseas investors to worry about then? Not if these figures can be trusted. But can they?
Kris Sayce, of Money Morning is even more of a sceptic on Australian housing than I am, and his sceptic’s eye spotted the fine print to this table: notice that there are 2 data sources, Demographia and UBS. A quick check of the Demographia data (check Schedule 1 on pages 31–37) shows that all the non-Australian numbers in that table come from Demographia, but none of the Australian numbers come from there. Kris then published a revised table where the Demographia numbers for Australia are substituted for the ones shown above–which by inference have to come from UBS. Kris observed that:
In order to make their point, the CBA have used the Demographia numbers as a reference point for all the non-Australian cities, yet they’ve used the UBS numbers for the Australian cities.
Why on earth would the bank do that?
Simply because if they’d used the Demographia numbers it would draw exactly the opposite conclusion to the argument they’re trying to make. The fact is, they’ve conveniently grabbed the bunch of numbers that fits their argument and discarded the ones that don’t.
If they’d used the Demographia numbers for all the cities, including the Australian cities, the table would look like this:
Paints a slightly different picture doesn’t it? Actually, it paints a completely different picture. One shows an unsustainable bubble, the other shows a bunch of figures comparable to elsewhere in the world. (Kris Sayce, Money Morning)
I suggest that any overseas investors who attend presentations by this team ask them the following questions about this table:
- Why did you use UBS data exclusively for the Australian cities, and Demographia data exclusively for the overseas cities?
- Demographia’s numbers show the “median house price divided by gross annual median household income” (p. 7). How are UBS’s numbers calculated?
- What would the ratios be for the non-Australian cities, if the UBS methodology was applied to them?
I would be very interested in hearing the answers to these questions.