Saturday, July 31, 2010

Busted in B.C.

A blog was uncovered at House Hunt Victoria that used statistical magic to indicate a seller's market was developing in Victoria. A Victoria Realtor calculated months of inventory* to be 3.5 in July even though it was a over 7.5 months in June. Not likely. It appears he uses a smaller set of data for inventory while adjusting a larger set of data for sales. House Hunt Victoria explains what is wrong with this figure:

It's really hard to follow because he's all over the place, but essentially yes; he's narrowed the current active listings down to a subset of SFH in certain municipalities and then used the market-wide sales data averaged over the first 6 months of the year, lumped pending and completed sales together to further confuse the data and then done the math to come up with 3.5 MOI but called it MSI (a term that also seems to come out of no where).

But I will be the first to admit I could be wrong in my interpretations somewhere as it is very difficult to follow his explanation/calculations because they are all over the place. I'd love it if he would clarify, but he seems to be too busy having his best year ever in 2010 to respond appropriately.
Anyway the Realtor posted and later deleted comments on HHV's blog. His incorrect post remains.

Well done HHV.

*Months of Inventory is the time required to sell total inventory at the monthly sales rate

Thursday, July 22, 2010

Calgary Census 2010

In the last post Radley commented on the low population growth in Calgary from the 2010 census.
Calgary’s population grew from 1,065,455 in April 2009 to 1,071,515 in April 2010. This represents an increase of 0.57%. This is the lowest percentage of growth experienced since 1984 when the city experienced an overall loss in population.

That's quite surprising considering the entire province had a rate of increase three times higher at 1.60%. See Alberta population report report for Q1 2010 I covered in this post.

The census has yearly statistics on the number of residential dwellings and population from which we can calculate a ratio to measure the amount of supply.

This chart shows population, total dwellings and the ratio of people per dwelling. This ratio was 2.54 in 1999 and fell sharply to 2.43 in 2004. The economy was doing well during this period letting Gen-Xers establish households. In addition to this there was low interest rates and the start of a housing optimism after many years of stagnation. I was surprised it was fairly steady throughout the boom years (2005-2007). During this time there was a lot of construction but also a lot of population growth.

This chart shows the change of population, and number of new dwellings each year. It looks like there was some reduction of supply in 2009 as population increased while construction dropped substantially. This reversed in 2010 as slow population growth lead to more supply on a relative basis.


Housing Analysis uncovered a Bank of Canada report on housing supply. It was full of incomprehensible formulas but had this comment on supply across the provinces:
Using aggregate data for Canada, our model suggests that the housing stock was overbuilt at the end of 2008 by 2 per cent. This gap is likely to close slowly by the end of 2018, given the projections of the key variables in the model. Provincial models indicate that there was likely overbuilding by the end of 2008 in Saskatchewan, New Brunswick, British Columbia, Ontario, and Quebec. On the other hand, the level of housing stock in Nova Scotia, Manitoba, and (surprisingly) Alberta fell back below what was consistent with fundamentals, after having risen above the level a couple of years earlier.

I wouldn't be as bold to say supply is below trend but it is difficult to make the case there is a large construction glut in Alberta. Having said that weakness in the housing market could persistent if migration remains low (or negative).

Saturday, July 17, 2010

19,000 sales? Unlikely.

From the June 2010 Realtors Association of Edmonton market report
“It has been quiet on the news front but very busy in REALTORS® offices as they list client’s properties for sale, book showings for buyers and attend open houses. This has not resulted in immediate sales, however, and, in anticipation that this slowdown will continue through the year, we have reduced our 2010 sales forecast by 2,000 units from 21,000 to just 19,000.”
“External influences pulled sales activity into the first four months of the year which reduced the demand in May and June. Overall there were 680 less residential sales in the first half of the year as compared to 2009,” said Westergard.

They have reduced their forecast to 19,000 based on how close we are to last year's sales. However last year started off in the middle of the financial crisis and sales continually increased from very low levels. 2010 has been the opposite of this. To show how unlikely this is see the level of sales that would be required for the rest of the year in the trend below (at a constant seasonally adjusted rate). Sales have to make an immediate and abrupt jump this month to a much higher rate and maintain that level going forward.

Looking at the trend I guess the year end total will be closer to 16,500 assuming a minor recovery in sales (shown below).

Friday, July 9, 2010

Another trip to the unicorn ranch

Come to the unicorn ranch and enjoy some refreshing Kool-Aid.

Good news! The 2009-10 Alberta deficit initially projected to be $4.7billion, later projected to be $6.9 billion, ended up being "only" $1billion. The deficit is more than offset by a $2 billion gain in heritage fund.
Note: there is still a hefty shortfall projected for this year

Have a great weekend.

Wednesday, July 7, 2010

Edmonton Market Statistics June 2010

The REALTORS Association of Edmonton released statistics for June and describe the market as "normal".

Instead of being normal sales were slightly worse than "scorched earth", a benchmark rate equivalent to the worst six months of the financial crisis.

Listings appear to have peaked early this year but remain somewhat high.

There are only 4 months in recent history which are as bad or worse in terms of seasonally adjusted sales - Nov 2008 to Jan 2009, when the world was falling apart and Sept 2007, following an insane boom.
Seasonally adjusted it is clear listings have eased up. Despite this inventory increased last month and remains very high. (9,406 active residential listings vs. 6,785 last year and 11,006 at the peak in May 2008)

Sales to new listing ratio was 41% seasonally adjusted. When this ratio dropped below 50% right after the end of the boom in 2007 and during the financial crisis there were corresponding price declines and this time is no different.

Initial explanation of seasonally adjusted data and benchmarks and seasonally adjusted sales to new listing ratio.

Saturday, July 3, 2010

Edmonton New Construction Update: May 2010

It's been three months since I have posted on the new construction situation in Edmonton. Since then single family home construction has increased and starts have remained high.

Single family home under construction peaked at 6,528 in August 2007 before falling to 1,764 April 2009 and finally rebounding to 3,760. It looks like the weakness in resale market has not yet had an impact on new homes. I expect to see price declines on existing resale inventory to lead a shift away from new homes as they become less competitive.

Total construction has been slower than single family alone. This is because condos took longer to recover with units under construction bottoming at 2,142 this March and have rebounded to 2,759 in May. While there is less oversupply going into this downturn compared to 2007 (10,314 vs. 17,781 in the chart above), weaker interprovincial migration and declining resale market will have a negative impact on new construction going forward.

Friday, July 2, 2010

Calgary Real Estate Statistics: June 2010

Sales are at "scorched earth" levels which is the same rate as during the worst six months of the financial crisis.

Listings are fairly high but down from the rate of March and April.

Seasonally adjusted sales to new listing ratio is 35%. A ratio of below 50% typically results in falling prices.

Raw data from Bob Truman's Old Criteria page.
Initial explanation of seasonally adjusted data and benchmarks and seasonally adjusted sales to new listing ratio.