The secret to house price rises

Yesterday I posted my observations that rates of credit issuance are the main driver for housing price adjustments in Australia. I noted that when the rate of credit issuance rose for a month then prices moved upwards soon after, and the reverse was true for the downside. It was therefore important as a housing investor and/or home buyer to take close notice of the trend in housing transactions to get an idea of where the market could be heading in the future.

I stated yesterday that my data was only for March 2002 onwards because that is when the ABS house price index data began. A reader mentioned that in fact there was also another dataset available so I decided to recheck my theory against this data.

The city “house price” charts below are simply  based on the data in the new dataset. The “credit churn” charts are created from the ABS state data for financial commitments for owner occupiers and the ABS state data for commercial finance commitments

The charts are as follows.

For Sydney:

And NSW credit churn:

For Melbourne:

And Vic credit churn:

For Brisbane:

And Qld credit churn:

And finally Perth:

And WA credit churn:

As far as I can tell in all markets above from 1991 until 2005 the rate of monthly issuance of credit was tightly coupled with changes in house prices. Again I am happy to be proved wrong. But given this data I would suggest that watching changes in housing market transactions, and therefore credit issuance, is an exceptionally good indicator of future house price movements.

Latest posts by __ADAM__ (see all)


      • Torchwood1979

        Happy days! And in regards to my comment last October:

        “Therefore I conclude we’re in a speculative bubble and when the penny drops with the flippers we’re in trouble nationwide.”

        Well the penny has dropped just recently, IMO. Now we’re in the Mexican standoff phase. Buyers aren’t willing to pay asking prices while too many sellers still have stars in their eyes. Either Government stimulus gets the buyers paying over the odds again or a bad turn in the economy starts the forced sales in large enough volumes to drive the market down.

          • Torchwood1979

            Yes, but we aren’t seeing distressed sales at a volume that will cause serious downward pressure – yet. With unemployment this low I believe the market will more or less hold. However what is overlooked by the “Low unemployment = safe RE market” brigade is the negative feedback loop caused by even small falls in RE values. Once those falls start feeding into the real economy and showing up in unemployment figures a correction or crash can start gaining pace. IMO we’re in the early stages of the unemployment upswing but it’ll take many months for it to really hit.

          • Understood, Torchwood.

            But the problem becomes, as i see it, when sales volumes drop and “discretionary sellers” (those who don’t HAVE to sell amidst a flat or declining market) begin to make up a smaller and smaller portion of the sales market listings.

            Ie. with time, the sellers AS A GROUP become increasingly “distressed” in their nature (proportion), as the discretionary sellers pull out.

            And, as the nature/fraction of the listed market becomes more and more distressed in charcater, prices will drop, as distressed sellers compete with distressed sellers – the phenomena is self-reinforcing…they just need to move their stuff.

            ie. it is BETTER (ironically!), price wise, for discretionary to STAY on the market, as their “stubborness” keeps the general level and nature of the market “elevated”; whereas distressed buyers are far more willing to compromise on price, as it is in their financial interest to do so!

            It is another reason why, IMHO, external shocks are not required – the system is capable of bringing itself down without outside “help”, so to speak.

            My 2c

          • …So, another point to be taken from all that is:

            It is not just the size of the remaining stock on the market that counts, but also its need-to-sell character.

            I’m sure there are metrics readily available that could well describe the relative need-to-sell characteristics at a given point of time; such that a time-serise of such metrics, alongside the level of sales listings (and other things) would provide an excellent base for robust analysis.

            eg. debt distribution (such as distribution of LVRs of sellers on the market with time)

            eg. time on the market distribution

            there would readily be others available, methinks…

          • Oh, and just to bang on a bit (because i do that, don’t i!).

            re: distressed sales, there is currently ~18,500 mortgagee sales on alone, as of today (there would be some double-ups there, but not many, as mortgagee sales tend to only go to a single agent/facilitator)

            (Yes, i track these sorts of housing figures! :P)

  1. Without seeing the plots of house price CHANGES and churn CHANGES on the same chart its impossible to see the correlation. The liner scales used also hide the large changes in churn in the 1990s when prices remain fairly flat.
    What I do see, for example, is:
    – a doubling of Victoria credit churn from 1992 to 1995 whilst Melbourne house prices stayed flat,
    – significant falling churn through 2000 in WA whilst Perth house prices actually rose.

    • I agree – Keen has shown the way on this subject for a while now.
      Time for a guest article from Mr Keen?
      I’d love to see it, if it focuses on AUS macro future 0-10 years out.

  2. I recently updated a Fairfax image on Bubblepedia which had made the same link between loans and house prices, although it’s already a little out of date after ABS revisions:

    Looks like after the FHB Boost prices divorced a bit from loan numbers, but are now looking to meet again at the new lower “normal”

  3. Why not look at mortgage credit spreads, over the ‘risk-free’ rate & compare them with house prices? I think you’ll see a good correlation, taking into account that houses are somewhat illiquid in that you can’t ‘day trade’ them.

  4. All good. Market should sort itself out, barring any more silly intervention from the muppets in Canberra. Here in Western Australia our economy is in better shape than the rest of the country yet house prices are slowing quite quickly.

    By the end of 2012 or 2013 i might become a property BULL. 🙂

  5. Here DE, you might be interested in this.

    It’s the AFG figures going back to pre-GFC. I have only had a very brief look but you can see the trending down in volumes of mortgages processed by AFG as housing credit growth has decellerated over that time. In some months back in the days of yore, they were processing 9-10 000 mortgages but now it’s down to a bit over half of that. The volume of mortgages is just as important a part of the housing credit growth story as prices.

  6. I had a chat to a RE bloke this morning. He has a friend works for Taylor Byrne property valuers. They are apparently flat out all over Australia at the moment doing property valuations for repossessed properties.

    • As i said just above, there is about 18,500 mortgagee (foreclosed, repossessed) sales on alone, as of today…the problem IS getting worse…

  7. Do they have a section that shows all these forclosures?? not that id buy one unless the rent yeild was towards 10% unlike the 3-4% now.

  8. That figures does not surprise me as I noted last month we had seen an upswing in the number of new loans mostly refinances few purchases. However that upswing is gone. Volumes are down strongly and management have sent a please explain.

  9. Are you saying that the May bounce reported by AFG appears to be fizzling out in June Stormboy?

    • Two ways, with one being better than the other:

      1) (Less good, for some reason): go to and do a search, but refine it with the keyword “mortgagee” and the Min Land Size to zero (“0”).

      This sort of search works for most of the major property listings websites.

      eg. for NSW: search for “NSW”, with min land size = 0, and keyword = “mortgagee”. As of today, you will get about 94 listings.


      2) Use google (etc) for a site-specific search of, to get,effectively, a nationalised result.

      eg. Do this google search:


      (to get this result URL: )

      You will get ~18,500 hits today.

      This is an indicative result from google 9as their will be double-ups, etc, agents ads, etc on the site also), but the discrepancy between the search types (1) and (2) makes me trust google search more, from an order of magnitude basis.

      As it stands taking the raw number of search (2) and the raw (state summed) total of of their sales listings (as can be found here:!australia), we can get an indicative idea of what fraction of residential properties on the market are “mortgagee” sales: (18500/425000)x100 = 4.35%.

      I think this is an overestimate.

      A modification of the google search in (2) with certain property numbers (looking for specific multiple listings, as a sample for over-representation, multiple listings) gives about half that halve multiple listings, and half that don’t (small sample size! so this is very quick at this point!).

      So, assume a double-up on listings for “mortgagee” via the google search on

      Hence, let’s halve the fractional finding: 4.35/2 ~ 2.18%.

      And then take some for general inaccuracy of the method.

      So, i’d say that, currently we could be looking at 1-2% of sales listings on the market today being “mortgagee” sales.

      Yes, i’m tracking this data, but on a relative, indicative level (eg. creating indexes; focrusing on relative movements, not absolute movements), not putting too much faith in the absolute numbers.

      Now you’ve got an insight into the BurbWatcher’s secret methods…! 😉


      Hope that made sense.


  10. Interestig correlation, but is there any causation there. Does one lead the other, and which way around?