Some data depicting monthly residential price changes is worthless: Terry Ryder

On Wednesday night I watched finance commentator Alan Kohler on ABC TV puzzling over the apparent contradiction of sharply-falling loans to investors and rising prices in the major cities.

His basic message was: it doesn’t make sense and it can’t continue. 

His basis problem was: it didn’t occur to him that some of the data he was presenting might be rubbish.

He shares this problem with commentators, economists and journalists right across Australia. No one ever questions the real estate data that’s pumped out by publicity-addicted research companies – not even when it’s patently absurd and clearly wrong.

We had the situation recently where one of big research firms published figures showing that house prices fell in six of the eight capital cities in the March Quarter. That was treated as gospel and interpreted as evidence that the alleged boom was over.

Around the same time another of the big numbers companies published figures claiming that house prices rose in six of the eight capital cities in the March Quarter. That result was reported as truth as well, with no one (apart from me, it seems) noticing that the two sets of figures totally contradicted one another and that two conflicting messages were being sent to the public. 

One source said Adelaide was leading the nation on price growth while another said its prices were falling at the fastest rate since 2012.

So back to Kohler who is very good at making sense of share-market events and economic data, but not so bright when it comes to real estate. He displayed a graph suggesting loans to investors have dived. He attributed this to the crackdown by APRA and the reaction of major lenders.

That was starkly at odds with reports in media only days earlier, claiming that “investors have begun flooding back into the market”. Newspapers reported that investors were 47.6% of all new mortgage commitments in March, the highest proportion since August 2015. This was attributed to “some lenders reversing the tighter lending requirements that were previously in place”. 

So Kohler’s graph told quite the opposite story to the figures published elsewhere. And he made the incorrect assumption – the Labor Party is making the same mistake with its negative gearing policy – that price rises are caused exclusively by investors.

His next problem - his confusion over prices rising so much against a background of apparently plummeting investor lending – was based on his blind acceptance of the May price growth figures from CoreLogic.

Like so many journalists and commentators, he fell into the trap of (a) accepting published figures as unchallengeable and (b) placing big emphasis on month-to-month movements. He’s been around long enough to know better. 

Much of the data purporting to depict monthly changes in prices in our major cities is worthless. Indeed, it’s worst than worthless, because it’s misleading and causes confusion for consumers. 

The CoreLogic figures for May would have us believe that the apartment prices rose 8.4 percent in Hobart. Seriously? Hobart units have added 8.4 percent in value in a single month? I visited Hobart just a week ago and there is nothing happening in that market to justify that figure, nor the one that says Hobart unit prices have risen almost 19 percent in the past 12 months. These figures are made of an artificial elastomer synthesized from petroleum byproducts (they’re rubbery).

The same research source claims that Darwin apartments prices fell 10 percent in the month of May; that Canberra apartment prices fell 6 percent; that Perth prices fell 6.3 percent; and that Sydney house prices rose 3.6 percent. All those dramatic movements in property values in a single month – are they believable? I would suggest not. 

Everyone reporting these numbers has expressed breathless amazement, but it hasn’t occurred to anyone that maybe the figures are a tad sus – or that short-term movements in median prices or value indexes are essentially irrelevant because a month is too short a time frame to deliver a meaningful pattern.

If you examine month-to-month movements in anything – retail spending, building approvals, the jobs market, you name it – you’ll find the numbers are quite erratic. Often they’re just statistical aberrations, but economists will be lining up to attribute great significance to them. 

Only the long-term trend matters and only that should be reported - and then only with care, because you can be sure that the numbers you’re recording today will be contradicted by another set from other source tomorrow.

Terry Ryder is the founder of hotspotting.com.au. You can email him or follow him on Twitter.