Forum Replies Created
Michael I’m afraid I don’t agree that statistics lie. They can either be massaged by statisticians, or the people interpreting them can missunderstand the bounds of their use.
On median prices I agree with you,and I would say that median prices are of less value because they dont decern between different land sizes, quality of housing etc. So if you have a situation where large block high quality housing was sold last quarter, and small block low quality housing was sold this quarter, then the figures will be unreliable.
Thats where I prefer the Residex model of pricing where you only compare a new sale to a previous sale of the same property. As long as you have an accurate information on any improvements theres little that can go wrong.
I’m not doubting that you will be right about which suburbs will outperform next, the fundamentals of the market are what drives it in the long term. But how can you believe that one area has always outperformed the others without refering to statistics?
Most importantly what can you determine if you know the statistics are accurate? For example I heard a radio story recently of an Australian who set up a statistical based horse betting business in Hong Kong and has made $250Million. He takes bets where the odds his stats gave a horse were way shorter than the bookies odds.
Remembering that the bookies set their odds based on the favour of the punters, the fact that he made so much money essentially shows the punters get it reliably wrong. I’m not saying this can be done in property, but I am saying that statistics dont lie.
Craig Sillitoe
Thanks Alistair, I’m sure you’re right but I wouldn’t call it a definitive answer. Sorry but I’m the type who just needs to see the figures, I find that way many assumtions are true but surprizingly useful ones are false.
Craig Sillitoe
hi debbrad, i may be a little late to this thread. what you are doing so far is quite cleaver. comparing only new sales to old sales of the same property. do you the annulalise the result, or better still quarterise it? have you put it in a spreadsheet, that would make it easier?
residex do a similar thing and have won awards because their formula is much more accurate than using median pricing. that would make it especially useful for small towns.
your data will be accurate no problem, the next question is how useful is accuracy in a small town thats likely to be affected more by random, unforeseeable events? what you need is to run some randomness testing on your data series.
the simplest form would be to check the ‘standard deviation’. this statistical study will tell you the how far prices have swung away from the average. you can then make an informed judgement about whether the price you are preparing to pay is low enough given the normal swing of prices in the area. standard deviation can be found under ‘studies’ in excel. though i use a share trading program and import property data into it.
there are further steps you can take along these lines if interested.
Craig Sillitoe
MichaelYardney i like your general premise that some suburbs are doing well in a down market. my question would be whats the best way to determine which suburbs will do well from now, should we assume that the ones that are up 20% will continue at a similr rate?
how else would one chose the right area in a choppy market? i’ve heard talk of sticking with ‘quality’ through tough times, but i’m yet to hear a definitive definition of what quality is and i’d love to see some data to back up the theory.
Craig Sillitoe
thanks derek, what you’ve said is pretty much where i was headed. cash flow status of an uncompleted deal needs to be judged based on the yield prevalant in an area not by the rent the tenant pays (if onen inherits a well paying tenant then its a bonus).
this leads me to the conclussion that (for cf+ investing) its all about the valuation (after any value adding). so the next question is what is the best way to value a property for cf+ investing? is it using yield, comparitives or median values? in my mind dividing the likely rent by the current median yield would be most accurate because median values and comparitives are more in line with the home buyer market than the rent market.
while i’m on the topic can i assume that in the current market an area with a ten year yield thats way above its current yield is therefore one that has been over bought (in investment terms), regardless of how high the yield still is? and could using yield be a way of judging when the tide is turning (as in yield curve)?
Craig Sillitoe
