sorry but i don't know how to c+p the graphs and pictures...so it might be beneficial if you read the original article...(it shows original evidence for 'rp data' and others...)
Why House Price Statistics Cant Be Trusted by Kris Sayce on January 6, 2010
Youre an idiot Sayce, dont you know the difference between Median and Mean Average?
That was just one reply to the Money Morning mailbag from reader GP following Mondays newsletter.
The email went on:
The three $20 million sales in Melbourne are virtually irrelevant as they are at one extreme whereas the Median is based on the value in the middle of the range. Get your facts right next time.
Yep, well take that one on the chin. Three sales of $20 million each they could be $60 million each, it doesnt matter much as long as its part of a statistically significant sample size shouldnt affect the median price to any great degree.
But what this slip-up proves is not so much that your editor cant grasp the definition of median although you can argue that if you like its that even the property spruikers cant figure out the best way to report the mega property bubble.
Weve taken another look at The Age article that we referenced. As clear as day the article states: The latest RP Data index, compiled from Valuer-Generals figures, indicates the Melbourne median house price hit a new record of about $580,000 in November.
The trouble is, when we refer to the RP Data press release itself, the number for Melbourne is just $486,400.
So theres obviously a discrepancy somewhere.
Perhaps The Age journalist made a couple of typos, and we should have checked our facts against the primary source.
Not so fast
Digging deeper into house prices Weve looked at the press releases from RP Data over the last year, and the reporting of the recent months figures strikes us as unusual.
Youll recall that weve scoffed at the Rismark International Hedonic Index. You remember the one, its that Einsteinian looking formula that magically shows property prices always going up.
Just in case youve forgotten it, here it is again:
Well never tire of publishing it. It makes us laugh almost as much as the Muppets on bikes photo!
The truth is, being a non-mathematician we wont trust any formula unless we can see a worked example. Its just how we are. Question everything. In fact, we wont even believe E=MC2 until we see numbers in there rather than letters.
Anyway, lets take a snapshot of the recent press release so you can see exactly how the numbers stack up:
You can see in black and white it states the Melbourne Median Price Over Last 3 Months is $486,400.
Stop there. Thats the first discrepancy. Last 3 Months what about the number specifically for November? Wheres that? Could that be the $580,000 number that The Age journalist was referring to?
Who knows, were too tight-fisted to hand over our own cash to a bunch of property spruikers so well have to guess.
But whatever the reason, thats a big difference [Tapping on calculator] almost $100,000.
The numbers still dont add up Then we thought of an ingenious idea Einsteinian almost. How about if we look at the median prices for the previous two months, crunch the numbers and see if we can figure out the median price for November.
So we did. Heres what the press release for Octobers numbers had to say:
Do you spot the difference? Apart from the fact the font is bigger.
Well, three things stand out. First is, shock horror, the index value for October is much higher than the average price for the three months ending in November. That could indicate prices actually fell in November, but lets not jump to conclusions just yet.
The second point is RP Data have produced a figure just for October. Not a three month median or average. Just October.
Finally, what happened to median? In November its there, but in October it isnt. Instead RP Data used Index Value.
What could it all mean? But its not fair to look at just two months, how about we go one month back from that to the September numbers:
Hmm, theres the median value. OK, this is getting too confusing. What about August
Right, well it states median values as well, so thats something. Oh, but now theres another fly in the ointment, because the heading states, RP Data-Rismark Hedonic Index Results.
But that also creates a problem for your pea-brained editor. Because were trying to figure out how you can realistically call something a Median price when youve used the following whacking great formula to make up a price:
As all the replies to Mondays Money Morning have told us paraphrased the Median price is like lining up all the prices in a row and then picking out the one that is right in the middle.
Which doesnt quite fit in with the fancy formula above. Unless you do a hedonic valuation on every single property based on sale prices and then calculate the median price.
But if thats the case, why would you bother? Surely the sale price is good enough sample size without crunching the numbers into a meaningless index.
Are you being duped on property prices? But what we really want to know is how is it possible for the Last 3 Month median price to be $486,400, when the October median is $509,987 and the September median is $499,794?
By our calculations Melbournes median house price would have fallen to $449,419 in order to achieve an approximate three-month median of $486,400.
Are you confused with all the numbers yet? We are. But that seems to be the way the property spruikers like it.
What puzzles us even more is the disappearance of the Hedonic Index from the RP Data press release. After all, this is the index that Rismark and RP Data have received accolades for:
RP Data-Rismarks monthly estimates are more timely and reliable than the ABSs quarterly readings. Rory Robertson, Macquarie
The RP Data-Rismark index has emerged as Australias authoritative source on home price trends. Craig James, CommSec
And from its own press releases:
Our innovative analysis and reporting tools are used by numerous industry bodies while government authorities use our very own indices and policy making.
But perhaps this explains it all. A note at the end of the most recent press release from RP Data. Its a long one so take a deep breath:
The median price is the 50th percentile observation based on all pooled home sales over the three months to end November 2009. This is different to the medians reported by other parties for several reasons. First, where appropriate it includes all property types (ie, not just detached houses, like the ABS). Second, the median value reported by the likes of APM is calculated using a stratification technique, which is different to the simple 50th percentile observation used here. RP Data-Rismarks previously reported median values must also be interpreted differently. These are the index values attributable to the RP Data-Rismark hedonic index, which was originally based at inception on median automated property valuation estimates (ie, the median of a statistical valuation of all capital city homes). The change in the index value over time reflects the underlying capital growth rates generated by residential property in the relevant region. These growth rates are not influenced by capital expenditure on homes, compositional changes in the types of properties being transacted, or variations in the type and quality of new homes manufactured over time. The RP Data-Rismark median values are not, therefore, the same as the simple median price associated with all homes sold during a given period. In future, we will report simple median prices to avoid any further confusion.
Ive underlined the key points.
Although incidentally, the stratification technique used by competitor APM which RP Data dont seem to like is also a method RP Data uses for its RP Data-Rismark Stratified Median Price Indices.
It seems the hedonic index which is more timely and reliable and an authoritative source and which is used by numerous industry bodies while government authorities use our very own indices and policy making has been replaced by simple median prices to avoid any further confusion.
Whats happened to the hedonic index? In a nutshell, it appears the hedonic index is being sent to the knackers yard. Instead they will now just line up all the numbers in a row and select the one in the middle. Even a median newsletter editor could do that.
That could make Rismarks four PhDs the most overqualified employees in the country.
Plus, all the previous figures disclosed by RP Data-Rismark are irrelevant in comparison because they used different methodologies.
So that got us thinking. Could it be that the reported $580,000 median house price for Melbourne was the number spewed out by the timely and reliable hedonic index?
Could it be that even the property spruikers realised there is something seriously wrong with an index that would have reported a gain of nearly 20% in the price of the average Melbourne house during the month of November?
Did they realise that no one in their right mind would have seriously believed such a price increase for average home values?
Could it be that the hedonic index was influenced by those three $20 million sales?
Do you know what? Weve got no idea what the answer is to any of those questions. Were too tight-fisted to write a cheque to these property spruikers in order to get the full report.
But it makes you think though doesnt it? What we do know is this
With most other investments you can get a pretty accurate idea of the price of your investment. For example, in the case of shares you can see second by second how much youre winning or how much youre losing.
But with property youve got absolutely no idea. All youve got to go on is the sale price of the neighbours house or dodgy statistics and indices invented by property spruikers.
Again, thats just the way the spruikers and the banks like it. The more ignorant you are about the real price of a house, the greater the chance is that youll believe the spruikers and take out the biggest loan you can to burden yourself with an unbearable debt.
But thats OK because property prices always go up.
Seriously, think about it logically. Does it really make sense to put down a $40,000 deposit so you can take out a $400,000 loan against an asset that you dont know the real value of?
Its mind boggling to see what the property spruikers can get away with. Ably supported by the limp-wristed mainstream press. We thought 2009 would be the year they would get exposed we were wrong.
So hows 2010 looking so far for the grand exposure? Keep reading and well see what happens.
"All content is - 2007 - 2009 Port Phillip Publishing Pty Ltd All Rights Reserved
We encourage you to republish our material, all we ask is that you provide a working text link back to the original article on this site. More information on our free financial content.
Port Phillip Publishing Pty Ltd holds an Australian Financial Services License: 323 988. "
