Sort by:   newest | oldest | most voted
CanuckDownUnder
Member
CanuckDownUnder

My biggest pet peeve regarding Vancouver housing blogs is people using the mean to track price changes. Please use median or even better a truncated median. If nobody has one handy send me all the sales data I’ll put one together for you.

With sales volume so low outliers can really move the mean around. Don’t worry about the noise.

Keeping An Eye On The Pimps
Guest
Keeping An Eye On The Pimps

Mike asks:

“Should I just give up on a correction, which in my mind, is long, long, overdue?”

Mike, no.

HipsterBear
Guest
HipsterBear

I like how tinfoil hat wearers shun the HPIs even though they are a better measure.

“As an investment banker I prefer medians and averages because those other validated models are too confusing” noone was heard to say.

registered
Member
registered

@3 HipsterBear: HPI is the ideal measure but as a ‘constructed metric’ it’s also very vulnerable to manipulation. Do independent third parties have complete access to the collection methods, raw data sets and methodology used in the MLS HPI? If not I see no reason to elevate them above RJ Reynolds medical studies.
Stick with the academically generated where possible, avoid those that derive the bulk of their funding from industries with skin in the game like Sauder. We really need a full Canadian equivalent of the Case Shiller.

yvr2zrh
Member
One stat that I track is a moving average based on a chronological list of sales. I usually track the last 100 transactions but we are so slow now that this is more than a month. So – if we do 80 sales, we will be on an approximate month basis. Here’s what we see. The current running average for past 80 units is: Average – $2,943,000 Median – $2,195,000 Looking back, this is very high for the average. There are 3 sales over 10 million in this so it’s pretty high. For the year, the lowest average was $2,200,000 which occurred in March. The highest average for the year was $3,058,000 occurred in February. As for Medians. The lowest was $1,824,000 which was in May/June. The highest median was also in February when the highest average occurred. So –… Read more »
watcher
Guest
watcher

give your head a shake on doubts regarding the new HPI. It was created for the real estate boards by academics with guidance from Statistics Canada and the Bank of Canada. All of the indicators are generated by industry, including the Case. The Teranet HPI is funded by the National Bank of Canada, and is created by a real estate services company.

watcher
Guest
watcher

S&P now owns the Case-Shiller index

HipsterBear
Guest
HipsterBear

@fixie guy: The MLS and Teranet track closely. You’re smart enough to run the numbers yourself to verify this.

Anonymous
Guest
Anonymous

I think the HPI tells the story. Regardless if it could be manipulated it always seems to track what is happening from my perspective. For example 2008 looked about right. Averages are not accurate at all. Inventory is a factor but way overrated on this blog. Watching daily numbers is silly.

Girlbear
Guest
Girlbear
29 Aug 2012 09:35 ET * Prices up 4.8 pct on year, 8th month of deceleration * Vancouver prices fell for the first time in five months * Further slowing possible in August By Andrea Hopkins TORONTO, Aug 29 (Reuters) – Canadian home prices rose in July from June to hit a record high for a third consecutive month, but the slower pace of gains and falling prices in Vancouver added to recent evidence the market is cooling. The Teranet-National Bank Composite House Price Index released on Wednesday showed overall prices climbed 0.7 percent in July from a month earlier. The index, which measures price changes for repeat sales of single-family homes, was up 4.8 percent from a year earlier. It was the eighth consecutive month the annual price gain decelerated. “Since prices rose 1.0 percent from July to August… Read more »
jesse
Member

@Anonymous: ” Inventory is a factor but way overrated on this blog. “

The data tell a different story. If you want best indication of where prices are headed in the short term, months of inventory and sell/list provide strong leading indicators. Supply/demand leads, prices lag.

watcher
Guest
watcher

@jesse – the ratios are coincident indicators, the magnitude of change is the leading

bearacha
Guest
bearacha

Can somebody “Rennie” these stats and chop off the 20% we don’t like?

Thanks

registered
Member
registered
@7 watcher Says:“All of the indicators are generated by industry….S&P now owns the Case-Shiller index” S&P is a primarily a ratings agency with minimal direct vested interest in the health of the real estate market. While by definition an ‘industry’, the difference between them and one that generates ratings on a product they represent, lobby and market should be obvious to children. Teranet is pretty good and the best we have but, partially tied as they are to the mortgage side, still not the ideal. Think the Berkley School of Economics predicting both the Dot Com and California real estate crashes. “…the new HPI. It was created for the real estate boards by academics…” Academics worked for big tobacco. The real estate industry’s were a terrific source of unintended humor in crashing markets everywhere so you might want to cite… Read more »
chris
Guest
chris

Correction in RE in Vancouver depends mostly in CAN YOU MAKE THE NEXT MONTHLY PAYMENT.

That depends on interest rates, unemployment rate. I am sure low interest rate is here to stay until end of the world. (Governments are already on the road to abandon the current currencies, you can cross the current in the currencies.) Unemployment factor is the sleeper here. (But any individual earnings salary is a sucker therefore precludes to this game of monopoly

Looking for a 20% drop/rise in a couple months? not likely.

VHB
Member
VHB

@fixie guy: I think an hedonic regression model is a good approach, so I’m inclined to like the HPI methodology. However, I do hold out some reservation that it hasn’t yet picked up the obvious changes in pricing in the market. If the MLS HPI doesn’t start picking this up in the next few months then I would have decreasing confidence in *this* implementation of the HPI model, although this wouldn’t diminish my belief that hedonic regressions are in general a solid approach.

I don’t buy the conspiracy theories at all. If the MLS HPI messes up, it will be because of poor modeling by some data nerds not because of dark conspiracies in smokey rooms.

patriotz
Member

@chris:
“Correction in RE in Vancouver depends mostly in CAN YOU MAKE THE NEXT MONTHLY PAYMENT.”

Depends on what “you” means. Current owners, no. Prospective owners, a big yes.

That’s what makes those mortgage changes so significant.

LazyCanadian
Guest
LazyCanadian

Article about falling profits due to more claims and less insurance being issued:
http://www.theglobeandmail.com/report-on-business/economy/housing/cmhc-sees-profit-fall-in-core-mortgage-insurance-business/article4507242/

The last paragraph caught my eye, “Banks like it because once those mortgages are insured, they can be securitized or sold into bonds.” I could take a guess, but I’m not really sure what this means.

jesse
Member

@watcher: ” the ratios are coincident indicators, the magnitude of change is the leading”

I cannot find much evidence of that, maybe I’m not looking at the right datasets. 2010 was a good example of where inventory growth and sales weakness in the first part of the year did not translate to weakness in the second half.

An interesting measure to look at more closely is average price changes leading HPI weakness. If the high-end of sales is the first to show weakness that would make sense. Ben Rabidoux had a good graph on that. And don’t anyone go using it without citation 🙂

jesse
Member
@VHB: “I don’t buy the conspiracy theories at all” The check I did is throw up mohican’s MOI-HPI correlation on Teranet and the MLS-HPI, both are highly negatively correlated (see here). If there were any fiddling with the numbers I expect that relationship would either lose correlation or become skewed. The best fit for MLS-HPI to MOI is with a 1 month delay. That is, if MOI is high in August, we won’t expect to see it until September’s MLS-HPI release. We’re already hearing of anecdotal sales going for less than assessed, and now in some cases less than previous sale price. The chickens are restless, but we should not discount that these specific anecdotes are not indicative of sales as a whole: they will tend to cause negative pressure but the rest of the market is perhaps not as… Read more »
Atomic Frog
Guest
Atomic Frog

I have got a lot of chinese friends who lives in Van West because those houses were owned by their parents/grand-parents completely mortgage-free.

The owners of these houses in these areas become inter-generational owner, they will just pass the house down to their kids/grand-kids

Anonymous
Guest
Anonymous

@Atomic Frog: So, then it’s a wash in terms of this phenomenon’s effect on the market: no added supply, no added demand.

jesse
Member
@fixie guy: The issue for me is not that the MLS-HPI is incorrect, it’s that the data are not transparent and held by specific people, who act as market makers and participants, for an advantage. I don’t know the reason MLS has put so much $ into the HPI, but one supposition, other than them wanting to provide some useful information to their customers, is that, if it is a valid measure of market movements, it provides a reason to keep their datasets private by showing their HPI is a reasonable gauge of general market movements. It also goes to show impartiality if it were to follow a peer-reviewed algorithm. I would think there is little to be gained to fiddle with the HPI given the pressure Realtor organizations face in most part because they are commissioned sell-siders who are… Read more »
HAM Solo
Guest
HAM Solo
@ Lazy The banks like getting their mortgage loans insured by the CMHC because it allows them to sell the mortgages (as “securities”) to institutions and other big investors for a higher price than they otherwise could. The reason the loans are worth more is because the credit risk of the borrower has been replaced by the credit risk of the Canadian government by the magic of CMHC insurance. Just to add to that, the paragraph you have focused on is the root cause of the insane Canadian housing bubble. Instead of the banks worrying about whether some Joey’s cocktail waitress can truly afford to carry an $800,000 mortgage on a Yaletown condo, they just originate the loans as quickly as they can and pass through the risk to the third party investors and the government insurer. The mortgages are… Read more »
Anonymous
Guest
Anonymous

“Should I just give up on a correction, which in my mind, is long, long, overdue?”

move on, do what best for you and your family. if you sit here and listen to these guys, your retirement would be just arriving as fast…
do you know that a few poor souls are still around here posting since 2006?

wpDiscuz