Thursday, May 08, 2008

CMHC Data - VHB makes an appearance

The VHB sent me some really interesting data this morning and I've spliced into my own data collection for our enjoyment.

Here we have the CMHC housing market data from 1948 until present. This includes starts, completions and units under construction. As you can see we are in completely unprecedented territory for the number of units under construction. Local builders really have no idea how much of a supply glut they are creating right now. This will be unfortunate for them but fortunate for me because I'd like to buy a new home at a steep discount. Bring on the supply!



As you can see from this second chart, population growth hasn't been the driver for this new construction and if we had more recent population data we would also see this trend continuing of a low ratio of starts to new persons. This will inevitably cause a supply glut and prices will fall. Builders are continuing to build because people are still buying the units they are putting onto the market.

The problem is that many of these people are either:
1) Speculators who plan to flip for profit - good luck with that
2) Wanna-be landlords who are going to bleed a tremendous amount of cash to hold on to their 'investment' and who will eventually sell because they can't afford to subsidize their renters, like me, for very long.

You can easily see that this is an unsustainable situation. Eventually the carrying costs of an average property must be equal to or less than the income from rent.

Thanks VHB. You give energy to this tired blogger.

22 comments:

mohican said...

This was interesting too http://tinyurl.com/6rpkfj

I'm looking forward to 50%+ price drops here in Vancouver.

Unknown said...

Interesting.

Looks like Zillow is way behind with the price adjustment for that street.

5865 is still estimated at $585,500

... also a ton of houses for sale in that area. Speculation can hurt...

http://tinyurl.com/6n4x8v

johnnyrent said...

Mohican

Probably two of the most telling graphs around.

I have a good friend who has made a fortune developing small condo and townhouse projects over the past six or seven years. He's now rushing to build the first component of a new townhouse complex on Oak Street. Suspect he'll have his first model ready to show sometime in the late Fall, with perhaps 25% of the entire complex complete at that time. I was thinking of sending him these graphs but I've thought better of it. These are dots I'm certain he wouldn't want to connect.

Regarding the Housing Starts/Person graph, numbers I have for Vancouver CMA for 2007 are population at 2,222,613 or a growth of 1.27% over 2006.

jesse said...

Great graphs, guys. I assume the second graph is actually "Net change in population per start"?

There was a guy commenting on vancouvercondo.info about how the 12 months of starts equaled the population growth for the next two years divided by the average people per dwelling so the amount of building is justified. Why is this logic wrong? Also the second graph does not acocunt for dwelling destruction.

jesse said...

Sorry that should be "24 months of starts". I think you get what I mean: the running total of starts seems justified based upon pop growth.

jesse said...

I think part of it is densification of SFH bungalows into larger sqft SFHs with basement suites, which would not show up in the construction starts/WIP/completions.

Van Housing Blogger said...

Nice work, as always, Mohican.

johnnyrent said...

Jesse

Starts only concern new construction, not existing stock of investor owned properties, nor properties that come on the market due a variety of reasons including transfers, retirement to other locales, illness, financial setbacks, divorce etc.

I don't know how acurate that 18,000 empty unit per BC Hydro figure that's been floating around is. Even if its off the mark by 30% however, its virtually enough to satisfy demand in the Vancouver CMA for a full year, nevermind new construction and existing units coming onstream for reasons previously mentioned.

johnnyrent said...

Sorry, I meant to say units under construction does not take into account existing......

mohican said...

Two additional comments:

1) We perhaps falsely assume population growth at current levels - this may not be the case in the future, especially if the construction sector contracts - hmmmm....

2) Large multifamily homes like the ones being built in Surrey / Langley and other burbs have enough space and the perfect setup for over 8 or more people to live comfortably. This would consist of 3 separate households. House / Carriage House / Basement Suite.

Van Housing Blogger said...

I just posted a nice graph with the starts vs pop growth over at RET. link.

the pope said...

Great work VHB & Mohican, thanks for putting these numbers together. If VHB doesn't mind you may want to add that absolute numbers graph from RET here - the divergence between population growth and construction in this boom is pretty amazing and that most recent VHB chart shows it very clearly.

happyrenter4life said...

Should there be an adjustment for declining household size? On the other hand, I don't think suites in SFHs show as part of the supply. Starts would therfore severely under estimate the number of actual dwellings since basement suites (or two) have become more common.

jesse said...

"Should there be an adjustment for declining household size?"

There is not much evidence of declining household size in the past 25 years according to statistics.

Van Housing Blogger said...

About household size. I dislike it when people project household size as a downward trend. There is a business cycle component. No job/less pay means the 20 something stays home longer, gets a room-mate, etc. This will hurt the condo market hard as the recession hits. Like the condo market needs more things to knock it down . . .

happyrenter4life said...

Jesse - You're right, no change in household size. Couldn't find the last 25 years, but here is the last 17:

http://www.gvrd.bc.ca/growth/keyfacts/perinprivhhd.htm

jesse said...

"No job/less pay means the 20 something stays home longer, gets a room-mate, etc. This will hurt the condo market hard as the recession hits."

If there were a housing shortage we would expect rents to increase as people densify (like London UK). But rents are flat, showing that dwelling formation is an "efficient market" and there is indeed more land being built on.

It's interesting to observe how rents in bubbly US cities are behaving as prices fall. I remember about a year ago some lobbyist for the local landlords' association was trying amass an effort for landlords to simultaneously raise rents. I'm not sure how that turned out.

M- said...

VHB and Mohican, great charts! I especially like the chart VHB posted on RET. That one shows how very clearly the housing starts have diverged from the population growth.

Maybe it's working for the time being, kids buying homes earlier than they normally would, because they know they're going to make money, but that trend works the opposite way when prices start to retreat-- people will live at home with their family, or room with others, rather than buy a depreciating asset.

patriotz said...

I remember about a year ago some lobbyist for the local landlords' association was trying amass an effort for landlords to simultaneously raise rents. I'm not sure how that turned out.

This is a conspiracy to fix prices and is, of course, a criminal offense. Although I'm not sure whether the line is crossed by just pumping the idea, or actually doing something.

Anyway it's impossible - cartels have trouble maintaining discipline even when they are legal and have only a few players like OPEC. Problem is that above-market prices always lead to oversupply, which provides an opportunity for windfalls for non-cartel members and an incentive for cheating by members.

jesse said...

"Anyway it's impossible"

I think it's impossible for a simpler reason than that: people densify when rents are too high -- the rental market is not as inelastic as people make it out to be.

jesse said...

Further to the "persons per occupied dwelling" front, here is a more detailed report from the 2006 census (PDF) and here is data from the 1996 census that has data going back to 1966. The number is trending down but interestingly bottomed in 1986 at 2.60 before going back up in the 1990s (2006 was at 2.58). I can certainly see the argument for smaller average dwelling size as cities densify but all it means is fewer people paying the rent.

It is a fair point that the number of rooms under construction is the important stat, if such a stat were available.

rentah said...

jesse: But rents are flat.

Yeah. If you look at the graph of real percentage per annum increase in rents for Vancouver, it's basically a bit of fluctuation around 0% to minus 1% for the last 10-20 years.

If this was the record of a company's earnings growth, what would one expect the stock price to have done over this period?