House price forecasts are usually wrong: HotSpotting's Terry Ryder

EXPERT OBSERVER

There’s a startling inconsistency in Australian news media when it comes to the treatment of real estate prices.

When actual data is published, showing what actually happened with house prices, media will savage the figures (if they are positive, as they were in March and April) and dismiss them as irrelevant or misleading. That’s journalism, right? Journalists are supposed to challenge things.

But if someone makes a forecast about prices which is negative, news media will accept that without question. And, worse, media will turn this speculation into a fact. It’s not a case of “prices are forecast to fall”; it’s presented as “prices will fall”.

This inconsistency in the approach of media speaks volumes about the mentality of journalists –

- If it’s a positive, attack it;

- If it’s a negative, make love to it.

Data from both CoreLogic and SQM Research showed prices resisted the impacts of virus shutdown in both March and April, with prices rising to some degree in most markets around the nation. We’re three months into the covid-19 schemozzle and there’s still no evidence of the price collapse predicted by the usual suspects.

But this was broadly challenged by news media as being misleading and untenable. Most headlines early in May, reporting the April figures, were extremely negative. One media outlet managed to turn all that solid price data into this headline: “Real Estate Markets Collapse.

It’s essentially about click-bait and the truth is optional – but, in other walks of life, if someone uses that level of deception and dishonesty to try to make money, they end up being charged with serious offences and in some cases go to jail.

Contrast all that with the treatment in the past week of forecasts for house prices by some of the big banks. These have been quite negative and news media has had a field day with those reports: no one has questioned or challenged them at all.

And they should be challenged, because the track records of bank economists in forecasting anything, but in particular real estate prices, are notoriously bad.

(As an aside, observe the record of Bill Evans, chief economist for Westpac, in predicting interest rates – he will make a forecast about likely movements in the next 12 months, with lots of publicity, but in perhaps two months’ time will issue a new, different, forecast. This begs the question: what was the value of your forecast two months ago, which you are now suggesting is wrong?)

If ever something deserved to be attacked by journalists, it’s the forecasts of bank economists about the housing market. Collectively, they have a very low level of credibility.

Let’s take National Australia Bank as a case study. The bank publishes regular forecasts about residential property prices, including the quarterly NAB Residential Property Survey reports. I’ve been reading NAB’s forecasts for many years and here’s what I know to be true – the forecasts are never correct. And I do mean never.

They always err on the side of pessimistic and invariably they are wrong.

Here’s a recent example. NAB published one of its reports in January 2019, containing forecasts for house prices for capital cities for the coming year.

Here what they predicted: capital cities on average would record a 3.8% fall in house prices in 2019. Sydney would drop 5.6%, Melbourne would decline 7.0%, Brisbane would deliver zero growth and other state capitals would deliver minimal growth.

News media, of course, gave this considerable fanfare and treated the forecasts as facts – i.e. Melbourne prices will fall 7% by the end of the year.

Here’s what actually happened, according to three major research entities which publish price statistics – and also according to NAB itself.

According to the latest NAB report, the capital city average for houses in 2019 was actually a rise of 2.9%, not a fall of 3.8%. The capital average for apartments was a rise of 3.4%, not the 4.3% decline that the NAB report forecast a year earlier.

The Melbourne result was a rise of 4.6% for houses and a rise of 6.5% for apartments, not the big declines that NAB had forecast. And in Sydney, there was a rise of 6.1% for houses and a rise of 3.4% for apartments, in contrast to the declines of 5.6% and 5.5% that the NAB report of January 2019 had forecast.

It’s quite a dismal failing, isn’t it, when you predict something will decline 5.6%, but it actually rises 6.1%.

NAB chief economist Alan Oster did not mention any of that in his recent media soundbites on his latest report. Here’s your credulity rating with me, Alan: 0

But the contrasts between the NAB forecasts and the eventual reality are even more stark when you look at the figures from more credible sources.

According to the figure from three major sources, the capital city average result for house prices in 2019 was a rise 5.5% (Domain), 2.1% (CoreLogic) and 2.9% (SQM Research).

The result for Sydney houses was a rise of 6.8% (Domain), 6.1% (CoreLogic) and 4.8% (SQM). Remember, NAB forecast a 5.6% decline for Sydney house prices.

The result for Melbourne houses was a rise of 8.7% (Domain), 4.6% (CoreLogic) and 5.7% (SQM) – not the 7.0% decline NAB had forecast.

There are similar contrasts with the other cities. Hobart, for example, rose 15.6% (Domain) and 11.7% (SQM), whereas the NAB report had forecast a 2019 rise of just 1.8%.

Without boring you with any more statistical detail, I can tell you that the NAB reports have been similarly, and consistently, wrong with their price forecasts year after year. They always lean towards negativity and real estate keeps on proving them wrong.

Yet, this nation’s gaggle of non-journalists has accepted their latest pessimistic outpouring without a single challenge. Its new forecasts, which are quite negative, have been celebrated in media across the country.

Shame on all of you for being so willing to heap further misery on Australian households without any credible justification.

Terry Ryder is the founder of hotspotting.com.au
ryder@hotspotting.com.au
twitter.com/hotspotting