Making Cents of Statistics

From: Steve McKnight


I noticed in the local Melbourne paper link: click here reported that Melbourne property prices rose 20% over the preceding 12 months (presumably Dec. 2000 to Dec. 2001). That's seems quite impressive!

But in some ways the way that it was reported in the paper was a little misleading. Not by what was said, but for what was left unexplained.

Almost hidden in the article is the word 'median'.

For people who don't know, 'median' is a statistical term that means the middle value in a data set. For example, in the data set being [21,22,23,24,25,26,27,28,29] the median would be 25.

In property, the median house prices are used instead of the average (or mean) house price because the average price can be adversely manipulated by unusually high or low prices. The median price is a much flatter trend since it is not skewed by value, just the number of sales in any given period.

The generally acceptable practice of taking the median house price and then extrapolating it across all properties (ie. from high priced inner city properties to the lower end junkers) is a statistical mistake.

It is only appropriate to apply the median house price movement to a property that was similar to the median property as defined in the previous analysis.

What am I saying? Be very careful about making assumptions that ALL property has increased by 20%. It is perfectly possible that property at the upper / lower end of the market moved a completely different percent or perhaps in the other direction all together.

I'm worried that people buying off the plan are coaxed into an expensive mistake based on the hype that ALL properties have risen in value. I can see the free wealth seminar where a copy of this article is handed out and the supposition that buying any kind of property will increase wealth overnight!

Another component of the article that caught my eye was that the average (rather than median!) amount borrowed was $160,329 with associated mortgage repayments of $1,044 per month.

Yet median (not average) weekly family incomes came to $988, or $4,278 per month, which is a debt:service ratio (used by banks to determine ability to service loan repayments) of (1044 / 4278) 24.4% - far less than the usual benchmark of 30%.

Perhaps this is an unremarkable kind of statistic except that I probably have too much time on my hands so I like to crunch different possibilities in my head.

For example, I'm interested to see what might happen in the event of a rise in interest rates.

Using my trusty financial calculator I worked out that the rate of interest implied to derive repayment of $1,044 per month is about 6.17%.

In order to 'max' people out in terms of getting them up to the 30% debt:service ratio, based on median monthly family income of $4,278, is 8.51%. You need to decide for yourself how realistic this rise will be, but interest rate rises seem to be back in the news.

My point in writing this article is to help you to understand why analysis based on median data cannot be accurately used across all elements of the property market.

I also wanted to highlight that when you are armed with a financial calculator, you can use the data presented to derive more information about likely opportunities in the future. I know that the housing market won't be quite as buoyant when interest rates increase to around the 8.5%.

As an investor I'd imagine that's the trigger point in my mind when the market swings from a 'seller's' to 'buyer's' advantage. Knowledge is power and now I have a mental trigger point to use to my advantage.

I hope you have found some use in this post.


Steve McKnight
Last edited by a moderator:
Reply: 1
From: Waverly Bay

Thanks Steve for the reminder on how to read median data and how to crunch more "meaning"/"signals" from seemingly basic data.

Personally, I find stats on affordability, median weekly household incomes, percentage of repayments over household income etc... very useful and interesting.

Alot of this info can be obtained from the AMP Banking/REIA Home Loan Affordability report (I get mine free . . .. but i think REI maybe gives it out for a charge).

For those interested, the report also calculates an "affordability indicator" (which is the ratio of family income to average loan repayments). Quite a reliable and useful statistic ....because it is calculated based on data from all the major lending institutions and is therefore a very comprehensive assessment of Australians' ability to afford their mortgage payments.


Last edited by a moderator:
Reply: 1.1
From: Alan Hill


May I ask how you get the Home Loan Affordability Reports for free? I'd love to do the same!

Last edited by a moderator: