Thursday, March 26, 2009

Teranet House Price Index - March 2009 - Monthly Report

A national downtrend, less pronounced in Montreal

Canadian home prices in January were down 2.4% from a year earlier, according to the Teranet–National Bank National Composite House Price Index™. As the chart below shows, this reading extends and deepens the home-price disinflation that began last February. It confirms that in early 2009, after more than five years of seller’s-market conditions, Canadian housing as a whole was a buyer’s market. January was also the fifth straight month in which the composite index was down from the month before, extending the first run of consecutive monthly declines since March 2007. The composite index is now down 5.5% from its peak of last August.

Of the six constituent city indices, three were down from a year earlier: Calgary (−8.2%), Vancouver (−4.2%) and Toronto (−2.4%). Two others were still up from a year earlier but by much less than last month: Ottawa (2.1%) and Halifax (1.2%). The sixth city, Montreal, also showed deceleration but maintained a respectable 12-month increase of 4.1%.

In another indication of the recent downtrend, each of the six city indices was down from its all-time high of last year (or 2007 for Calgary). However, they have not been moving in lockstep. The Calgary index was down from the month before in 14 of the last 17 months, with seven straight declines from last July through January. January was also a seventh consecutive month of decline for Vancouver. For Toronto it was the fifth, for Ottawa the third and for Halifax the second (the fourth in five months). Montreal’s index peaked in September, fell for three months in a row, then edged up 0.1% in January.

The historical data of the Teranet–National Bank House Price Index™ is available at http://www.housepriceindex.ca/.

Calgary -11.4% from peak

Halifax -3.5% from peak

Montreal -0.4% from peak

Ottawa -3.6% from peak

Toronto -6.1% from peak

Vancouver -8.3% from peak

National Composite -5.5% from peak.

By : Marc Pinsonneault, Senior Economist Economic & Strategy Team, National Bank Financial Group

11 comments:

AndrewJ said...

This seems lower than I thought for the decline in Vancouver. Was the ~15 percent number being thrown around based on the benchmark price?

mohican said...

Yes, the REBGV benchmark price has fallen nearly 15% so far.

It is possible that the repeat sales methodology lags a bit in showing price drops or the REBGV benchmark is off a bit, or both.

patriotz said...

If someone can show me a repeat sale from spring 2008 and winter 2008/2009 that has had less than a 10% drop, I'd sure like to see it.

Because all the examples I've seen have been a lot more than that.

JimTan said...

"Because all the examples I've seen have been a lot more than that."

Why don't you prepare your own index?

jesse said...

I thought a 3mo moving average was used in this index but correct me if this is not so. That could explain the lag compared to the benchmark.

patriotz said...

Thanks Jesse, that certainly could. I believe the Case-Shiller has a similar lag.

Why don't you prepare your own index?

Well maybe because it would be of no real use to me beyond the data we already have and I have better things to do?

If you want to challenge someone's observations, show us some data, rather then making smart remarks. That's what this forum is for.

jesse said...

Yeah I confirmed the filtering here (PDF). They use a 3 month weighted filter in the regression coefficient domain.

They are transparent with their method (though not the data) but to really understand it one pretty much needs a degree in statistics.

John Collison said...

Well, according to this data, there's really no problem. An 8% drop in Vancouver from peak?!?

Okay, then. Bust? What bust?

Move along, nothing to see here.

An 8% price drop in the midst of a global economic and real estate meltdown is actually as good as an INCREASE.

There is absolutely something FUCT with these numbers. Even in Calgary, I know people with houses selling 25% below peak prices.

What gives?

JimTan said...

"If you want to challenge someone's observations, show us some data, rather then making smart remarks. That's what this forum is for."

Exactly! If you don't like the teranet moving average numbers, show better numbers.

mohican said...

I don't think there is anything wrong with the numbers at all. The results have more to do with statistical methodology than with what has actually occurred. The months ahead should show further price drops in this index.

I agree with patriotz that we would be hard pressed to find a repeat sale with less than a 10% price drop.

patriotz said...

Exactly! If you don't like the teranet moving average numbers, show better numbers.

There have been a whole lot of repeat sale numbers on Real Estate Talks, a forum I think you're familiar with.

An index is just a statistical representation. The issue is not whether one "likes" the numbers, or whether another index would be "better". The issue is if the index is seen to lagging actual sales information, why is this happening and does it represent some fault or is it just a consequence of the methodology which will correct over time. It would appear to be the latter.

But don't go telling a poster that if he has issues with the index, he should start his own. That's sort of like telling someone that if he didn't like his flight, he should start his own airline.