Forecasting real estate myths: Pete Wargent
Pete WargentDecember 17, 2020
Last week Reserve Bank Governor Stevens disclosed that he had discussed with his underlings whether the Reserve should do away with its apparently precise central growth, unemployment and inflation forecasts, in favour of publishing only their 'fan charts' and ranges in tables.
By way of an example the November Statement on Monetary Policy (SOMP) forecasts showed GDP growth returning to 3 to 4% by the year ended December 2017, with the unemployment rate forecast to decline gradually, but the fan charts show that even the 70 per cent confidence intervals are surprisingly wide.
Noting that accurate forecasting is impossible, Stevens lamented that even if even if the central lines were to be left off the fan charts altogether, those awkward commentator folk would simply back it of the charts, which seems rather unfair.
In short, the most powerful central bank or econometric modelling can't forecast accurately what will happen over any meaningful time horizon, and neither can you!
In short, the most powerful central bank or econometric modelling can't forecast accurately what will happen over any meaningful time horizon, and neither can you!
Ponzi scheme
Also last week a macro market report hit the headlines likening the Australian property market to a Ponzi scheme. Australians would be forgiven for thinking that they've heard such warnings before, for the simple reason that they have, more or less continuously since 2001.
I haven't read the report in question, so am in no position to critique the content, but to his credit Lindsay David of LF Economics has previously made predictions that are specific, measurable and time-bound, and so his predictions can be judged accordingly.
I haven't read the report in question, so am in no position to critique the content, but to his credit Lindsay David of LF Economics has previously made predictions that are specific, measurable and time-bound, and so his predictions can be judged accordingly.
Co-author Philip Soos has nothing if not longevity, rolling out his crash predictions for such a long period of time that they are threatening to see off their fifth Prime Ministerial term of office (his views were expressed in some detail in the piece "Bubbling Over: The End of Australia's $2 Trillion Housing Party" which predicted a 40 per cent crash for housing values nationally).
The last time anyone bothered to check the value of Australian dwelling stock had swelled to $5.8 trillion with the latest available data implying that $6 trillion is a shoo-in, probably before the end of the calendar year.
To be fair, on the one hand dwelling prices have indeed corrected sharply in any number of mining towns and regions: Gladstone, Port Hedland, South Hedland, Dysart, and Moranbah spring to mind, and no doubt there are plenty of others.
On the other hand figures released by Residex at the end of last week showed that Sydney's median house price has increased from $668,000 at the end of 2010 to $1,058,500 by the end of October 2015, an increase of more than 58 per cent or nearly $400,000.
Buried under the emotive rhetoric Soos has made a number of robust arguments, in particular the illusory nature of housing shortages, which are seemingly apparent when the economy is humming along, but can wondrously vanish into thin air when market downturns happen.
These simple points highlight the perennial challenges facing macro property market research reports and forecasts: market timing and granularity.
On the other hand figures released by Residex at the end of last week showed that Sydney's median house price has increased from $668,000 at the end of 2010 to $1,058,500 by the end of October 2015, an increase of more than 58 per cent or nearly $400,000.
Buried under the emotive rhetoric Soos has made a number of robust arguments, in particular the illusory nature of housing shortages, which are seemingly apparent when the economy is humming along, but can wondrously vanish into thin air when market downturns happen.
These simple points highlight the perennial challenges facing macro property market research reports and forecasts: market timing and granularity.
Timing
While macro research can identify downside risks and probabilities, history suggests that it is very difficult to forecast housing markets accurately over any meaningul time horizon.
Famously Case & Shiller called the US real estate bubble in July 2003 when its index read 142.99, yet the index ran all the way to 206.52 in July 2006 before reversing all the way back from whence it came to 134.07 in March 2012 (the latest Case-Shiller reading in 2015 showed the index to be 28 percent above the level it was tracking at the time of the original bubble call, while homeowners have continued to benefit from housing services throughout).
There would be similar tales in Northern Ireland where average house prices all but tripled from £88,000 to £226,000 between 2002 and 2007 before crashing all the way back to £125,000. An obvious bubble, sure, but the timing was difficult (if not impossible) to pick. The average price today has increased again to £162,000, approaching double the average house price in 2002.
In London we have heard crash predictions for at least the past decade-and-a-half, while despite downturns and recessions house prices in some boroughs have tripled...and still there has been no crash.
Few would disagree that parts of Australia's housing market are seriously expensive, but timing the downturns is rarely as easy as it appears in hindisght.
Granularity
The state level data shows that since September 2011, New South Wales (+57 per cent) and Victoria (+23 per cent) have accounted for 80 per cent of the increase in the total value of dwelling stock, while the figures for the June 2015 quarter showed that prices are declining in Western Australia and the Northern Territory.
Click to enlarge
The obvious point here being that there is no "Australian property market", rather a series of sub-markets which cover 9.6 million dwellings and demonstrate very different characteristics and fundamentals.
Having observed the history of bubble predictions Scott Sumner once concluded that:
"If we notice market movements that seem to align with our initial forecasts we tend to pat ourselves on the back and assume the forecasts were correct. This is one of many cognitive biases that human beings are prone to. Pay no attention to bubble forecasts. They are useless. Indeed the entire bubble concept is useless."
Having observed the history of bubble predictions Scott Sumner once concluded that:
"If we notice market movements that seem to align with our initial forecasts we tend to pat ourselves on the back and assume the forecasts were correct. This is one of many cognitive biases that human beings are prone to. Pay no attention to bubble forecasts. They are useless. Indeed the entire bubble concept is useless."
Risk markets
Last year Soos wrote that due to it having the strongest yields Darwin has the smallest property bubble, a theory which might hold some sway in an academic vacuum, but shows a worrying lack of understanding of the real world and what caused the Top End's housing boom in the first place.
In fact, in my opinion at any rate, Darwin appears to be the capital city with by far the greatest risk of a genuine price crash, with rents now in freefall and vacancy rates rising sharply (cf. the illusory supply shortage identified by Soos himself).
Thin property markets such as Darwin with a relatively small population are more prone to shifts in dynamics at the margin, and as population growth switches into reverse gear by the end of this calendar year the results to the downside could be spectacular.
Other high risk markets include any number of regional areas with a residual resources capex footprint, with mining capex now collapsing, while parts of inner city Melbourne and Brisbane are suffering from overbuilding of new apartments, with APRA's lending crackdown set to leave many off-the-plan buyers short.
Even within capital city markets there are sub-markets which can be expected to experience diverging fortunes.
For example, with the population still expanding rapidly, Sydney has a number of inner city markets for which demand is so high that any correction in prices will be met by a wave of new buyers. On the other hand, price inflation in some of the outer western Sydney locations over the past few years has been so far beyond absurd that even a 40 per cent correction might represent a neutral scenario.
Property Observer recently reported on the report by LF Economics comparing Australia's housing market to a Ponzi scheme, which you can read here.
PETE WARGENT is the co-founder of AllenWargent property buyers (London, Sydney) and a best-selling author and blogger.
His latest book is Four Green Houses and a Red Hotel.
Pete Wargent
Pete Wargent is the co-founder of BuyersBuyers.com.au, offering affordable homebuying assistance to all Australians, and a best-selling author and blogger.
Tags:
Property market