« Why are forecasters predicting a mild recession for Canada? | Main | The Bank of Canada's Financial System Review »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Hmmm. Short-term labour flows increase the demand for housing services? (You move to Alberta for work, and rent an apartment there, but don't sell the house in New Brunswick, where the wife and kids are still living?) Just a guess. If I'm right, vacancy rates should rise, and rents fall, with the decline in the Alberta oil fields and construction markets, as the workers move back home.

Maybe. But the trend appears to be nation-wide.

How about people not willing or able to get mortgages? Demand moves from owner occupied to rental.

It's not like the supply of people goes down in an economic downturn (modulo migration).

If people were unable to get mortgages, and stopped buying houses, then I can see two possibilities:
1. Other people are unable to sell houses, and so stay in their houses rather than renting (which would leave the vacancy rates on apartments unchanged).
2. Empty houses sit unsold, and stay empty (so the total vacancy rate of the housing stock stays the same, but more houses are empty and fewer apartments are vacant).

I'm trying to think clearly about this:

Change in number of vacant houses plus apartments = # of new units built - number of new households formed + number of households "doubling up" (moving back in with the parents) - number of households "undoubling" (moving out of the parents' basement).

Demand for houses is down because getting credit is much more difficult and houses are too expensive relative to incomes. If people can't get a mortgages or simply don't want to pay too much for a house (especially when they are worried about losing their jobs), then they become renters. This increases demand for rentals so vacancies go down and rents go up. After all, we all have to live somewhere.

I am wondering will credit contraction and unemployment crush demand and will foreclosures increase as people who paid too much during the bubble start to loose their jobs. I dunno, but it seems like a distinct possibility.

I believe the same trend was observed in many American cities, in the early stages of their decline.

Don't know about the nationwide trend, but cities like Vancouver have a lot of potential rentals that could hit the market - as prices were increasing steadily, condo investors had an incentive to leave units vacant - especially new units, which could then be sold in like-new condition. A lot of these people weren't interested in landlording, they were after the capital gain. They may well change their minds with declines in prices. And if they sell, the next owner might well be interested in renting the unit out, especially if these things finally start to cash-flow again. (Right now, even those who rent out the units typically see a monthly loss.)

It wouldn't surprise me if we followed the same path as many American cities, and saw a softening in rents over the next little while.

And, as in the above comments, household formation patterns can change drastically with a change in economic fortune: people take in roommates, move back home, move out later, etc.

You're forgetting a term for obsolete housing (condemned, destroyed etc.) unless you meant net new units built.

Andrew: yes, it should have been net new houses built.

Patrick: If demand for houses goes down, what happens to the houses those people would have bought but didn't? Don't they get rented out? If so, the supply of rentals increases by the same amount as the demand. Or do they sit vacant? If so, do they show up in vacancy data?

For what it is worth, find the latest and maybe best Canadian House price index.


It uses historical data one each house to determine the increase in Price for Canadian houses in 6 market (Toronto, Calgary Montreal etc etc)

[edited to make the click-through link - SG]

NLF: That is worth quite a lot! I was browsing through the site, thinking "Oh well, it's easy to read, but nothing special, not many cities, ho hum...", then I clicked on the methodology section:

"Properties with at least two sales are required in the calculations. Such a “sales pair” measures the increase or decrease of the property value in the period between the sales in a linear fashion. The fundamental assumption of the constant level quality of each property makes possible the index calculation but imposes difficulties in selecting (or filtering) those properties that satisfy it."

That's important. It's immune (or at least much less vulnerable) to the composition bias problem which looks at average sales prices over time.

How does it compare (in method) to Case/Schiller?
Is it new? Does anybody know about it?

Wow. That's a great resource; I foresee many hours working through that. Thanks so much!

Which reminds me; a few months ago I stumbled across this site http://www.homevaluationscanada.com/ (how do you make those links work Stephen?) which will give you an estimate of the value of your home (or anyone's house) for $30 in a few seconds, over the net, based purely on the previous selling price(s) of your home, and looking at comparable houses. So just out of curiosity I paid my $30. The estimate it gave me seemed about right. But it was based on only one previous price for my home, which is what I paid for it 10 years ago. (They knew what I had paid for it and when; I didn't tell them.)

I was trying to figure out the econometrics. My guess is that it's some sort of fixed effects model, and they try to estimate the change in value of my house based on the change in values of other houses which have sold two or more times in the same neighbourhood, and which had about the same estimated value as my house did when I bought it. Now, I had it appraised for a mortgage, and the appraised value came in 10% higher than the price I paid (it was a foreclosure sale, so I bought it "sans aucune guarantee"). My guess is that if I substituted the appraised value into the econometric package, it would also spit out a current value around 10% higher as well.

Anyway, getting to the point: my guess is that the site NLF links us to is probably using a very similar econometric method to the one I link to above. Maybe it's the same people?

You're the econometrician Stephen! Good topic for a post. I think there's a very big difference between the National Bank estimates and the CRA (?) estimates, IIRC. I think TD Bank did a rough and ready estimate to get rid of the provincial composition effects, but National seems to be well ahead on this one!

There is a liquidity component to this as well. A house with sitting tenants is MUCH harder to sell. People will leave houses empty if they can in order to get the price they want. Some of the houses on the market are empty, if the average time on the market goes up, the number of empty houses goes up.

Nick: they'll show up as unsold inventory, or prices will crash to the point where people will buy them, I suppose.

Looking at the graph for Calgary at the link you provided, it certainly looks eerily like the case-schiller graphs did about 6 months ago...

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad