Meddling with Medians
When researching a property or a suburb, one of the research items everyone should consider is the median price. There is only one problem with this though: how do we know the median price quoted is accurate?
You can be forgiven for always believing the median, after all, quoted medians are based on heavy research by well respected companies, but be warned, not everything is as it seems.
Not everything is as it seems
Here is just one example where you may be caught out. It is the quoted median price for houses in an inner suburb of Brisbane for the year ending June 2008
Year ending June 2008
12 Months Growth
* YIP = Your Investment Magazine
* API = Australian Property Investor Magazine
Both companies are quoting median prices for the same product, in the same suburb, over the same period of time, yet the difference between the quoted median prices is $42,500 (or 7.7%) and the difference in growth a whopping 118%!
It is hard to imagine that there can be such a giant difference, particularly given both magazines use data supplied by two of Australia’s leading property researchers: RPData and Australian Property Monitors.
Keep in mind what the 'median' actually is
So how do we explain the difference? Firstly we must remember what we are looking at: the ‘median price’. The median price is not to be confused with ‘average price’. The median price is the middle sale, or put another way, if there were 21 applicable sales and you lined them up from lowest to highest, the median sale would be the one in the middle, sale number 11.
Knowing this, you should never look at a median price where only a small number of sales are presented. There have been many headlines over the years along the lines of ‘Growth increases by 18% in the June quarter’, yet when you look a little closer there were only 4 or 5 sales in that quarter. In theory, the higher the number of sales being compared, the more accurate the median price should be.
All this said, each magazine quoted a reasonable number of sales, being 64 for YIP and 82 for API, so you would expect a small difference, but nothing to this magnitude.
Another thing to be remembered is that some suburbs can vary greatly in price within itself. The inner city suburb of this example has quite a number of houses on small lots (approx 300m2 each). It also has quite a few properties fronting busy roads and near commercial or industrial uses. Then there pockets with larger lots, beautiful houses, city views and handy locations. In short, there is a wide variance in house and land qualities within the suburb and therefore a wide variance in prices.
How the median price can be inaccurate
To demonstrate this point, we’ll use a hypothetical example where we have a suburb with 11 sales. Of these 11 sales, 6 are of houses on small lots or fronting arterial roads. In this example we’ll say that they all sold for $350k each. Then you have 5 houses that sold on the only hill in the suburb. They are larger houses, larger lots and have city views. The sale price of these we’ll say was $650k each. The median price for the suburb is the middle figure, which would be $350k in this example. Yet the average for the suburb would be $486k, a very large difference. Obviously this is an extreme case, but believe it or not it happens.
Michael Matusik from Matusik Property Insights releases a Snapshot each year called ‘24’. This is a Snapshot of 24 suburbs that have been ‘cleaned’ of extraordinary sales, such as divorce sales, where one party sells 'half' of the property to the other . Further, Matusik also analyses property resale where no substantial improvements to the dwellings have been made. In other words, he is looking at directly comparable property and their growth as a result of market growth alone, not because of improvements as a result of renovations, developments, or similar. This last method is interesting, because there is no doubt that in certain growth markets and certain suburbs, you can have many sales after properties have been substantially improved. What Matusik found in this year’s ‘24’ was a remarkably large difference between the median price of property as traditionally viewed and the median price of his ‘cleaned’ data.
Unfortunately, as Matusik points out, it is cost prohibitive to use his method, hence the reason he only uses this method on 24 suburbs each year, so in the meantime we are stuck with the traditional approach.
Michael Matusik is presenting at Developing for Profit, the ultimate two day workshop for new and experienced developers. For more information on this workshop click here. Note that it is only a month away!
With this in mind we go back to our original example of the inner city Brisbance suburb to see why there is such a large difference between the quoted medians.
Interestingly in this example, if we jump back 12 months to June 2007 by removing the quoted growth of 17% (YIP) and 7.8% (API), we arrive at a much more aligned median price of $506,410 and $510,204 respectively. The strange thing here is when you review the quoted figure from API for June 2007, they actually quote $515,000 and not $510,204 as has been determined based on their stated growth to June 2008.
So on this basis one could assume that it comes back to the sales captured by each in that 12 month period – those 18 extra sales quoted by API must mainly be at the lower end of the market.
BUT, just when we think we are getting to the bottom of things we decide to go back and look at what each publication has quoted in the months leading up to June 2008 and we find one very disturbing anomaly. The figures quoted for Queensland by YIP for June 2008 are identical to those for YIP for the year ending April 2008! We suspect this is an error on their part where they released the June figures a little ahead of schedule!
Yet these are the figures relied upon by thousands of investors each year and quoted in the media.
Is it any wonder we are having trouble finding sense with the world financial markets at the moment?
Without knowing the exact sales that each company has used we cannot possibly determine who is more accurate, so what do we do?
Meddle with the medians
There are four suggested approaches, each with their own positives and negatives:
1. Get a third opinion from another respected company such as Residex and run with the middle figure.
(John Edwards from Residex will be presenting at the special 'Crisis Response' edition of Property to the Max to be held on November 1st and 2nd in Brisbane. Click here for more information.)
2. Average out the median prices collected.
3. Go with your gut instinct, which can be surprisingly reliable if you are researching an area you know.
4. Do your own research.
Take into account the stated timeframe
One other major factor that you should be well aware of when viewing median prices is that they are the median for a given period, most often quarterly or annually. The positive with the shorter time frame is they more accurately portray what is happening here and now. The big negative is that there are less sales, particularly if you break it down to smaller areas such as suburbs and therefore the data becomes less reliable. This issue is somewhat mitigated when we look at an annual figure, however there is one big negative here as well in that it captures sales for a lengthy period and as a result it may not reflect very accurately current market trends. For example there may have been a total of 200 sales over the course of 12 months, 150 of these may have occurred in the first 6 months during a growth period and the remaining 50 sales occurred over the remaining 6 months during a quiet growth period, yet the data will still state good growth, even if that growth only occurred in the first half of the figures.
It is highly recommended you take the time to confirm the data you receive and where possible, consult with someone with experience in the area that you are looking at buying in.
Investigate Property are buyer’s agents in the Brisbane area, which means their full time job is to look at investment property and analyse and research every deal. This ensures that their clients are not paying too much for a deal and that they are buying properties with multiple opportunities.
For more information on the benefits of being a client click here.