The real word on Melbourne house prices

ScreenHunter_02 Apr. 09 08.45

Cross-posted from David Collyer at Prosper

‘A Guide to Property Values 2012’ by Victoria’s Valuer General was released July 5.

It has been comprehensively ignored.

The report dissects property transfers and prices lodged with the titles office. This dataset is the actuals, the definitive historical record, and is what scholars of the future will use to examine The Great Australian Land Bubble.

There are substantial delays between sale date to lodgement to collation. Records are provided to the Valuer General thirty days after settlement, which is typically sixty to ninety days after price is agreed and then aggregated and released quarterly. Five per cent of 2012 sales still remain unrecorded in June 2013. I have condemned this time gap elsewhere.

Let us look at what has actually happened.

“The property market in Victoria showed an overall downward trend during 2012”, says Robert Marsh Victorian Valuer General.

ScreenHunter_25 Jul. 17 07.35

Median house prices peaked in 2010 ($420,000) and 2011 ($420,000). They have since fallen – yes, fallen – to $412,000 in 2012 and $370,000 in the ‘snapshot’ of the first few weeks of 2013.

The Victoria-wide data is blurry and disguises as much as it reveals. The municipal data is clearer, and still statistically valid. I follow Boroondara, Maroondah and Wyndham.

Alert readers will know I rent in leafy Balwyn North, a 4 bed 2 bath double garage house worth $1.2 million (my estimate) when I moved in 2010. The rent is $600pw, which the owner did not seek to increase on renewal. The owner is making a 2.5pc gross return. Net rental return would be less than half that, say, 1 pc.

Balwyn North is part of Boroondara municipality, a very affluent area inhabited by professionals and business owners. Overall education levels are very high and local schools, both private and government, consistently rank among the best in the state. Transport links and community facilities are fully developed and excellent.

Median property prices peaked at $1.32 million in 2010. They are now $1.15 million, down $170,000 or 14.7 pc. The 2013 datapoint comprising 111 house sales is statistically valid.

ScreenHunter_26 Jul. 17 07.37

Maroondah, on Melbourne’s Outer East is a developed dormitory area. Eastward expansion is constrained by the Urban Growth Boundary. Educational levels are state-average, schools are good, transport is adequate though it is a long commute to the CBD, community facilities are good due to the early foresight of local leaders but now declining.

Median prices peaked in 2010 at $510,000. They are now at $470,000, a decline of $40,000 or 8.5 pc. The 2013 datapoint comprising 157 house sales is statistically valid.

ScreenHunter_27 Jul. 17 07.38

Wyndham, on Melbourne’s South West is a new, rapidly changing dormitory area adjoining a heavy industry zone and without physical limits to the west and north. Education levels are below average; schools, community facilities and transport links are comprehensively overwhelmed by rapid population growth and are in serious infrastructure deficit.

Median prices peaked later, at $359,000 in 2011. They are now at $340,000, down $19,000 or 5.5 pc. The 2013 datapoint comprising 387 house sales is statistically valid.

ScreenHunter_28 Jul. 17 07.39

Vacant land, limited in Boroondara and Maroondah, is plentiful and actively offered in Wyndham. Median land prices were $220,000 in 2011 and are now $202,750. This figure probably understates the price movement as developers seeking to maintain margins have likely reduced the size of newly subdivided lots as well as restricting new supply in the face of weak demand. Wyndham land sales volumes peaked at 6,338 in 2009 and were a mere 1,383 in 2012.

The Valuer General was kind enough to supply a 25 year price history, which includes the last land price downturn in 1989-91. That downturn was mainly a commercial price retreat. Some juicy highlights of that commercial bust can be found here.

In 1989, Boroondara median house prices were $250,000 and fell 13.1 pc to $221,000 in 1991. Maroondah had a longer downturn, prices fell 10.9 pc from the 1989 median peak of $132,000 to $119,000 in 1995. Maroondah next matched the 1989 peak a long nine years later in 1998. Wyndham house medians topped in 1990 at $107,500 and retreated 7.5 pc to $100,000 in 1995. Inflation and interest rates were much higher then and the real price change – particularly in Maroondah – were more damaging that these bare figures suggest.

So, the current price retreat almost exactly matches the 1989-91 experience. In my opinion, the land price downturn has barely begun. Australians are saturated in debt. Land prices are still at record levels in relation to incomes. Successive interest rate cuts are having less and less impact. The key valuation tool, rent to price, is drunk on its own success. Please, do your own assessment, draw your own conclusions.

Leith van Onselen


  1. What is the degree of variation, if nay, between these figures and the Rismarks et. al. of the marketplace?

  2. Excellent work, now how does my council justify an increase in my property value for rate purposes?
    The entire property data scene is rigged.

    • Agree, property values based upon real estate agency official valuers, upon which local councils rely for valuations and thus rate income, while state governments have been addicted to stamp duty and federal sees real estate as the industry (and related) to be major employers of Australians and replace mining……

      The report probably confirms what many have suspected (in Melbourne at least), versus those who are in denial (cognitive dissonance), after encouraging their friends etc. to become burdened with mortgages as property always goes up… you will miss out…… population growth…… cheaper than renting etc. etc.

  3. “….Please, do your own assessment, draw your own conclusions……”

    Excellent advice, to go with a model example of what we “need to know”. Great work, Collyer.

  4. reusachtigeMEMBER

    Of course, we won’t see this reproduced in the msm – it doesn’t show the picture they would like us to see. Wrong data.

    • thomickersMEMBER

      MSM can still use spin the charts to show that housing doubles every 7-10 years (except for the end bit lol) 😛

    • Why not? It clearly shown that we’ve already hit the bottom by comparing with the previous cycle and its the best time to buy before the next twenty years of boom time, isn’t it?

  5. flyingfoxMEMBER

    Great data! So I wasn’t imagining things. I remember visiting Melbourne in 2006 and thought housing was quite cheap/competitive compared to Sydney. When I moved here last year, it had gone through the roof. Mining investment boom ?

    If we are quickly approaching 2008 levels, I wonder when the investors will start feeling the pinch…

  6. Rumplestatskin

    That’s a cool report.

    One point (which is not bad per se, but needs to be noted).

    The big problem with using median/mean of actual sales that the composition changes over time? Much of both the size of the boom and bust on this measure could be a product of the timing of sales of high end property.

    For example, if I want to sell a multi-million dollar home, I wait until boom times. But if I am selling at the low end of the market a) I am more likely to be forced to sell because of financial difficulties, and b) the timing probably has less impact on prices than at the top end. Therefore this measure could amplify the price cycle.

    Of course, the flip side is that this measure DOES show that homes are transacting at lower prices, which is very useful complementary information to the hedonic index approach.

    The other very important information in the report is the declining sales volumes. So not only are lower priced home transacting, fewer of them are. This really shows the behavioural responses to constrained credit conditions.

    It would be nice to compile the valuer-general’s historical data across all State to provide a nice interactive database….

    • Is data available on the total value of house transactions x time interval? That would control for variations in the composition of houses being sold and also variations in the numbers being sold.

      • It wouldn’t control for variations in the composition of houses being sold; to do that you would need to track actual houses over time and try and work out the proportions of sales that are occurring in each quartile of the STOCK rather than “the market” at different points in time.

        I think using the “median” is the best “blunt instrument” statistical smoothing tool, which is why “median multiples” are so highly regarded by some authorities.

        A factor I would very much like to see disaggregated at all times, Rumplestatskin, is the fluctuation in the LAND VALUE per square foot. This is where MOST of the fluctuation occurs. This would tell us a LOT.

        It would also be easier to control for variations in the locations of houses being sold, as the value of land can be tracked “by location” much more simply than “the value of houses” by “quartile”.

        Under conditions of urban growth containment, the value of land per square foot trends inexorably upwards to take out any slack in the “price of housing” that might be occurring by households reducing their consumption of land. This is why 1/10 of an acre sections in the unaffordable cities are several times the price of 1/4 acre sections in an affordable city. But around the rising trend, very high cyclical volatility will be found in the unaffordable cities.

  7. “Vacant land, limited in Boroondara and Maroondah, is plentiful and actively offered in Wyndham. Median land prices were $220,000 in 2011 and are now $202,750. This figure probably understates the price movement as developers seeking to maintain margins have likely reduced the size of newly subdivided lots as well as restricting new supply in the face of weak demand.”

    Unsure what you mean by understating the figure the reduction in lot sizes would be overstating a reduction in median land prices not understating it.

    • The % decrease in price per sqm will be higher than the % decrease in the total lot price if the av lot size is decreasing.

      • The volatility in the price per sq m will be many times higher than the volatility in “house prices”.

        The run-up will be colossal, because section sizes have been sacrificed even as the prices per section have risen.

  8. Fascinating report.

    Is this showing that even with deleveraging barely begun house prices are already sliding down. What does this mean for the banks?

    Btw, nth Balwyn is a nice area – I grew up there. But I thought you owned a house somewhere in the eastern suburbs?

  9. thomickersMEMBER

    Hi david,

    Are you noticing the knockdown re-builds (French box-style homes) emerging in the Balwyn area?

    Based on what I’m seeing from the ground level, the vacancy rates for these builds (6 months after completion) are about 50% ($2-$3million sale price wanted and most not even listed/taken down)

    Ghost homes perhaps?

    • Yes, thomickers. Boroondara is being mock-French Provincialled. Next to me is a brand new mock-mock-mock 4 car & theater basement, four living, five bed six bath confection. Exterior walls are rendered polystyrene already moving and cracking. Ornamentation is as credible as a lipsticked rhino.
      No allowance was made for a clothes line so the exterior full-kitchen, 8 burner BBQ and splash pool spa entertainment complex is draped with drying clothes every weekend by the new owners. They paid $3.5m for that privilege.

      Real end-stage bubble stuff.

      The accellerated activity in Boroondara is builders and speculators constructing to exit their land holdings as quickly as practicable. The exhausted 50-60 year old houses being replaced are unrentable, so owners are doubling down and praying they can sell before things get too much worse. They will need to be quick.

      Don’t Buy Now!

      • reusachtigeMEMBER

        Sounds like Strathfield in Sydney. It’s the same knock-down rebuild mentality, and English does seem to be a second language. Here too you can rent an older house worth $1.2mil for $600/wk. Paying a mortgage off at 95% of that value would cost you more than 3 times as much. Madness.

      • This is the thing, i don’t think mum and dad blue and pale-blue speculators quite understand just how much white collar executive australia has geared up. I know guys under forty looking at holding more than 2mil in pure residential debt and guys in their forties easily up to 3-5 mil. This is pure cost of debt cash-flow stuff – the balance sheet can be what ever you want to believe (until you can’t).

        That’s a big bad bet there.

  10. Does this mean market forces now are working in favour of lower prices without undue government intervention….much like the AUD…

    At this rate may well negate the argument in the other post here today!

    • You need to massage your message 2D. The real risk is the mining bogans will realise they have been cheated of their boom by tax, land prices and bad government. Your employers are at risk of being strung from a lamp post by angry men driving leased lime green SS utes.

      Time to switch your call to Law and Order. Your salary depends on it.

      • I don’t work for the government. I work for the sector that almost single handedly saved the economy 😉

      • GunnamattaMEMBER

        But I see you have taken to freelancing for the Real Estate lobby of late.

        And that sector you work for has done a great job for the last five years or so – but it would be taking a couple of points off GDP growth per year heading forward?

      • Holy shit!

        Mining bot when did you come out of the PR mine (lather, rinse and repeat, – lather, rinse and repeat) with your Canary?

        “I don’t work for the government. I work for the sector that almost single handedly saved the economy ;)”


        🙂 *grin*

    • GunnamattaMEMBER

      Well when you think about how much government intervention and RBA positioning has gone into propping them up……..

      The issue isnt about additional government intervention – it is about doing something about the intervention already there.

      Although for some of us it is also the chance to make ageing pro real estate relatives in full lecture mode perspire a a tad.

    • It may correct on its own but not before it has caused years of carnage for folks that only want the certainty that comes from owning your own pad at a reasonable price.

      The goal here is to get dopy folk to expose their balance sheet to the finance industry. The fallout has been debt slavery or worse rental no-mans land for many families – particularly those that are not savy white collar finance workers that are happy to take the little landlord risk.

      • actually he works for the European and US aristocrats that own the footloose capital that winds through the jurisdictions to the ownership of the mining houses. They are in partnership with the Chinese gov for now.

  11. Volumes of sales is the other untold story of the report. 2012 was the lowest number of House sales in Melbourne since 1995, lowest number of units sales in Melbourne since 1996 and lowest number of sales of Vacant Residential land since the series began in 1985.

    Multiplying the mean prices by the volumes gives a dollar figure 26% lower than the peak and lower than 2008.

    • I should have added the last time volumes fell for houses for two years in a row was in 1989 & 1990.

    • reusachtigeMEMBER

      No one wants to sell if they think things will get better soon, unless they have to sell. That’s when things will really change.

    • The Patrician

      “lowest number of sales of Vacant Residential land since the series began in 1985.”

      Thanks Cornflakes. That is a telling stat. Land prices per sqm through the roof and sales at all time lows. A clear indication that the price mechanism is broken. No wonder those swelling landbanks are so full.

      Hey DC, when can we expect an update on your landbank register?

      • reusachtigeMEMBER

        Only in real estate can such a price mechanism exist. I’d love to see land bankers go bust, severely.

      • The unit volume is the more remarkable.

        There were 22,824 in Melbourne units sold in 2012 (plus 3,374 in Country Victoria).

        From today’s building completion data, it can be seen there were 19,277 units completed in 2012.

        Given that some of the sales would have to be older classical flats (I live in a 1960’s six-pack that has two sales in the past year), there must be next to no market in re-sale of the modern apartments being built.

  12. This is fantastic reporting and analysis, I always enjoy reading UE’s articles. I can never find this kind of intelligence in the Mainstream, I guess it’s too unpopular to be the bearer of bad news.

    What I find interesting, taken from the Valuer Generals statistics, is the large changes in CBD land values that have occurred in Melbourne’s past, These indeed do coincide with the asset bubble of the 80’s and following corporate cycles and downturns of 93′ and so forth.

    It’s amazing that they doubled every year from 1988 to 1990, then sunk afterward.

    Currently the inner-city property prices do look like they have peaked, yet they look to be demolishing the Palace Theatre, which is an Icon.

  13. I concur. Recently bought a home in east melbourne susburb which would provide a 2.5% gross yield. Mad without question. If we actually had a housing “market” in Aus then there would undoubtedly be a crash. But the reality is we dont – it is govt controlled via land restriction policies, concessions to nimbyism, high immigration, tax incentives, low rates and money printing, fhogs etc etc etc. I bought becasue I am sick of renting and because I have become not so certain than govt intervetion will not continue to keep prices elvated for my lifetime. Its a gamble, and even if house prices remain elevated I doubt I will come out on top given massive rent v buy annual cost discrepency. …. blame my wife

  14. What would be interesting is to compare this dataset with dataset provided by other providers to see if there are any discrepancies…