Statistical Deception

Hotspotting

The misuse of price statistics represents a clear and present danger for real estate consumers trying to make choices about where to buy.

The much-quoted adage about lies, damned lies and statistics applies very aptly to median prices for locations across Australia.

While this data can be useful to buyers and sellers, if used intelligently, way too often it’s misused and abused in news media in ways that misinform and mislead consumers.

One of the most common misuses of median price data occurs when media outlets publish lists of the locations which, allegedly, have had the biggest growth in property values in a recent time period.

Journalists love these lists, usually spat out of the computer database of a research organisation which craves free publicity and doesn’t care too much about the accuracy or authenticity of the figures.

One of the problems is that journalists often confuse a 10% rise in the median house price with a 10% rise in the location’s property values. Often it’s not the same thing at all, because median prices are very rubbery figures.

Here are a few facts about median prices you need to know about:-

1. If you do a computer search on the median price for any suburb or town in Australia, you might get answers from seven or eight different sources and they will be all different.

2. If you ask how much the median house price has grown, or fallen, in the past 12 months, you again will often get seven or eight different answers.

3. Median prices are notoriously and dangerously unreliable if the sales sample is small. If, for example, there have been only nine or ten sales in a suburb or town in the past year, then the median price will be meaningless, and the increase or decrease will be unreliable, because that’s a very small sales sample.

4. At Hotspotting, we disregard median price data for a location if there are fewer than 30 sales in a year.

So recently, a recent media headline shouted very loudly about a New South Wales location where “property values” had risen 150% in the past five years – including 8.2% in the past 12 months, according to CoreLogic – which is one of those research organisations which loves free publicity and doesn’t always scrutinise the data that achieves it.

The reality is that the location in question, Catherine Hill Bay in the Lake Macquarie area, is a very small village with very few sales – and the figures on its median house price cannot be treated as gospel.

According to the article, the median house price was $1.43 million, according to CoreLogic, up 8.2% in 12 months and 151% in five years.

But if you check out the latest figures on yourinvestmentpropertymag.com.au, the median house price is $1.56 million, up 5.7% in the past 12 months – and has grown at a rate of 25% per year over the past 10 years – which means property values are doubling every three years.

If that was true, this insignificant location would be the outstanding real estate performer in the nation, if not the world.

But PropTrack’s latest information says the median house price is $1.6 million, up 10% in the past 12 months. But with little increase in the past two years.

But here’s the thing. How many house sales in Catherine Hill Bay in the past year?

Just 10. Which means the median house price data is rubbish.

If you look at the PropTrack graph for the change in its median house price over the past five years, the figures jump all over the place – because there are so few sales.

The message is: if you torture statistics enough, they’ll tell you anything you want to hear.

But smart investors will not base a big purchase decision on this kind of data.