Sunday, April 06, 2008

REBGV Chart Extravaganza

This was a lot of work so I'm going to take a few days break and let this data sink in for a bit. Cheers. Thanks to Paulb and realtylink.org for the data. The charts show data from the Real Estate Board of Greater Vancouver including sales, listings and more.


This last chart is interesting because it indicates that we should see negative price changes once inventory reaches the 5.5 to 6 months level.

21 comments:

Anonymous said...
This comment has been removed by a blog administrator.
Strataman said...

Terrific Mohican! You deserve a break! Thanks for all the hard work!

van_coffee said...

Out of curiosity, what is the coefficient of correlation and the R-Squared value on your last chart?

Thanks,

fish10 said...

Great graphs.

Many thanks.

Craig said...

Interesting trend for the sales/listings graph. March 2008 was the first month in years to decline rather than rise from February.

This perhaps more than anything suggests it will be different this year.

mohican said...

craig - I agree with your conclusion. Feb to Mar has not seen a rise in Months of Inventory for the past few years and indicates that perhaps things will change this year.

The coefficient of correlation between Months of Inventory and Quarterly Price changes is -0.7969 and the R sq value is 0.635.

BobbyBear said...

Nice work Mohican.

Similar trading range of 12 to 22 weeks that happened in the U.S. for several years before sharply increasing. Of course as MOI increased dramatically the pressure on prices intensified.

The MOI chart at CalculatedRisk was a smoother chart probably because of the huge number of transactions done in the U.S.

A very strong start for listings this year. The winds of change are a blowin'?

Magic number 27 weeks of MOI?

van_coffee said...

Hi Mohican -
I fogot to thank you for putting this together. I don't often post on your site, but am a regular "lurker".

Is there much difference on the correlation coefficient in the monthly data vs the quarterly data? I would expect it to be modestly lower (absolutely), but should still be close.

An R squared of 0.64 is pretty high for a single explantory variable, and given the stationary natre of the data series, it is actually probably a legitimate R squared (lots of times these are spurious and the result of using non-stationary data).

Anyway - good work, and it's great to see some old school "VHB Style" graphs posted.

VC

mightymouse said...

Wow! Thanks Mo!

I hadn't realised how much things had really changed this year until I saw your charts!

Etienne de Cochon said...

Thanks Mohican. Great work!

wizardofozziejurock said...

Wow, nice indeed Mohican!

Our local REIC might want to consider upping their spin dosage - some people are starting to figure this out.

AndrewJ said...

A question about Paul's numbers.

How come the overall inventory is going up more than the list minus sell number? It seems very strange.

johnnyrent said...

Rentingsucks

Don't know about Paul's numbers overall, but on April 7th there were 351 listings and 85 sales, a difference of 266. Total inventory increased less than this, however, at 185.

AndrewJ said...

I usually chalk differences in that direction up to expiries and it has always happened. This is a differences in the other direction.

I can understand a little bit of discrepancy but for the last week it seems huge.

April 1st to April 7th numbers:

New listings: 749
Sales: 263

Total invetory increase should be 486.

Actual increase looking at the inventory numbers is 740. That's 254 extra listings. Is inventory that out of sync with the sell/list numbers?

AndrewJ said...

I seem to remember the sell/list numbers our out of step with each other by about 2 weeks. If that's the case then the sale numbers in 2weeks time are going to be very, very bad to make up this difference assuming the inventory numbers and listings are correct and up to date.

Paul said...

Renting sucks:

Everything is correct. Often properties come back on market or expire,get cancelled or get terminated. That can account for the daily numbers not adding up easily. Often listings expire at midnight and I post the data after 9pm.

Paul said...

Sales usually lag about 2 weeks.

freako said...

Great stuff Mohican.

I agree with your conclusion. Feb to Mar has not seen a rise in Months of Inventory for the past few years and indicates that perhaps things will change this year.

Unless this is a false start (we've had a few) we are about to see the mother of all inventory buildups. So far it has an uncanny resemblance to the inventory buildup that occured in Phoenix, San Diego, Las Vegas etc in early 2006. The year over year sales trend will be very telling.

Sales usually lag about 2 weeks.

As the market slows, could we not expect this to increase as fewer buyers make unconditional offers?

tulip-Mania2 said...

Tick Tock, Tick Tock

Mark Fenger said...

Paul

"Often properties come back on market or expire,get cancelled or get terminated."

I'd just arrived at that conclusion when I scrolled down to your answer. But it does raise an interesting point. If enough sales are falling through due to lack of financing or whatever to show up on the charts it could mean that things are already getting pretty ugly.

AndrewJ said...

Actually it isn't anywhere near as bad as I said above. I messed up and was using only the SFH numbers and not the condo numbers and then comparing them to the total inventory. If you add the condo and SFH number together you get the usual pattern of inventory not quite increasing as much as the list - sell. Sorry about that guys.

Thanks for the numbers Paul they are very much appreciated.