Housing Myths and
Misconceptions:
Vancouver and Toronto
May, 2015
Helmut Pastrick
Chief Economist
Central 1 Credit Union
Housing Myth #1:
 Canada’s housing market is overvalued…..
in particular, Vancouver and Toronto
Some commentary on housing prices:
“Overvaluation is especially marked in Canada,
particularly with respect to rents (89%) but also in
relation to income (34%).”, The Economist, Apr. 16, 2015
“…houses appear overvalued but prices are still rising.
This is the case in Canada …”, OECD. (overvaluation:
30% by income, 66% by rents)
“Fitch deems … BC to be overvalued in real terms by
26%...”, Canadian Residential Mortgage Loan Loss
Model, May 2014.
Some commentary on housing prices:
“homes in Canada are 63 per cent overvalued, ….
Values in Canada are 35 per cent higher when the
median house price is compared to the median
household income than the historical average and 91 per
cent higher compared with average rentals.”
Deutsche Bank AG, Jan. 2015
….and others too numerous to list
The Bank of Canada on overvaluation in
Canadian housing markets:
 Bank of Canada study concluded housing prices
‘could be’ 10% to 30% overvalued
 Study examined 43 house price cycles in 18 OECD
countries from 1975 to the present
 Model included national data on real house price,
real per capita disposable income, and a long-term
(10-year) interest rate
Source: International House Price Cycles, Monetary Policy and Risk Premiums, Bank of
Canada Working Paper 2014-54, December 2014, pp 23.
Source: Financial System Review, Bank of Canada, December 2014, pp 17.
On overvaluation in Canadian housing markets:
 Bank of Canada model does not include supply side,
only two demand variables; not stable over time
 Supply factors such as land supply availability,
development and new construction costs, and land
use regulations also determine house prices
 Housing markets are local in nature; overvaluation
concerns in Canada are almost exclusively about
Vancouver and Toronto
 Assessing the degree of house price “overvaluation”is
a complex exercise requiring assessment of both
demand and supply factors
Long-term uptrend in housing prices
MLS Residential Average Sale Prices: B.C.
Dollars – thousands
600
500
400
Current $
2002 $
300
200
100
0
1961
1971
1981
1991
Source: CREA, Statistics Canada, Central 1 Credit Union.
2001
2011
Latest: 2014
Long-term uptrend in housing prices
MLS Residential Average Sale Prices: Toronto REB
Dollars – thousands
600
500
400
Current $
2002 $
300
200
100
0
1966
1976
1986
1996
Source: TREB, Statistics Canada, Central 1 Credit Union.
2006
Latest: 2014
Growing divergence between average and
median sale prices
Residential Sale Prices: Metro Vancouver
Thousands
Dollars - thousands
800
700
600
500
Average
Median
400
300
200
100
0
1976
1981
1986
Source: Landcor Data Corp.
1991
1996
2001
2006
2011
Latest: 2014
Land makes up most of total property price
Single Detached Property Price Components: Metro Vancouver
Thousands
Dollars - thousands
800
700
600
Total
Land
Improvements
500
400
300
200
100
0
1990
1995
2000
2005
2010
Source: Landcor Data Corp. Note: Based on assessed values. Latest: 2011
Little increase in Metro Vancouver land area
last 30 years
Land Area: Selected Metropolitan Areas
Square km.
7,000
6,000
5,000
1981
2011
4,000
3,000
2,000
1,000
0
Calgary
Toronto
Source: Statistics Canada Census.
Victoria
Vancouver
Less than one-half of metro land area available
for urban development
Population to grow by nearly 1 million persons in
next 25 years
Metro Vancouver Population, Actual and Forecast
Persons
3,500,000
3,000,000
2,500,000
2,000,000
1,500,000
1,000,000
500,000
0
1981
2011
Source: Statistics Canada Census, BC Stats forecast 2036.
2036
Increasing land utilization
Population Density, Actual and Forecast: Metro Vancouver
Persons per square km.
1,200
1,000
800
600
400
200
1976
1986
1996
2006
2016
2026
Source: Statistics Canada, BC Stats population forecasts, C1CU. Latest actual: 2011
2036
Higher housing and land prices in the long term
Metro Vancouver Population Density and Housing Price
Persons per square km.
Dollars - thousands
1,200
1,000
1,000
800
800
600
600
400
400
200
200
0
1976 1986 1996 2006 2016 2026 2036
Source: Statistics Canada, C1CU forecasts. Latest actual: 2011
Density (L)
Actual $
Projected $
Higher housing and land prices in the long term
Household Density and Housing Price: Toronto CMA
Households per sq. km.
Dollars - thousands
600
1,000
500
800
400
600
300
400
200
200
100
0
1976 1986 1996 2006 2016 2026 2036
Source: Statistics Canada, C1CU forecasts. Latest actual: 2011
Density (L)
Actual $
Projected $
Price-to-income ratio well above long-term
average; sign of overvaluation and a bubble?
Residential Price-to-Income Ratio, Metro Vancouver
Ratio – sale price to non-elderly family income
7
6
5
Actual
Average
4
3
2
1
1980
1985
1990
1995
2000
Source: Statistics Canada, Landcor Data Corp, C1CU.
2005
2010
Latest: 2014
Housing affordability to worsen in long term
Residential Price-to-Income Ratios: Vancouver and Toronto
Ratio – average sale price to average family income
13
11
Vancouver
Toronto
9
7
5
3
1
1976
1986
1996
2006
2016
Source: Statistics Canada, CREA, Central 1 Credit Union.
2026
2036
About 25 years ago …ratio average less than 3.0
B.C. Residential Price-to-Income Ratio, 1961 to 1991
Ratio – average sale price to average family income
4.5
4.0
3.5
Actual
Average
3.0
2.5
2.0
1.5
1.0
1961
1966
1971
1976
Source: Statistics Canada, CREA, C1CU.
1981
1986
1991
Long term average varies with time period
B.C. Residential Price-to-Income Ratio, 1961 to 2014
Ratio – average sale price to average family income
9
8
7
Actual
Av. 61-14
Av. 80-14
6
5
4
3
2
1
1961
1971
1981
1991
2001
2011
Source: Statistics Canada, CREA, Central 1 Credit Union. Latest: 2014
In 25 years, yesterday’s overvaluation becomes
tomorrow’s undervaluation
B.C. Residential Price-to-Income Ratio, 1961 to 2041
Ratio – average sale price to average family income
11
9
7
5
3
1
1961 1971 1981 1991 2001 2011 2021 2031 2041
Source: Statistics Canada, CREA, Central 1 Credit Union.
Actual
Av. 61-91
Av. 80-14
Av. 80-41
Av. 11-41
All Average
Accepted definition of an asset bubble:
“[I]f the reason that the price is high today is only
because investors believe that the selling price is high
tomorrow – when ‘fundamental’ factors do not seem to
justify such a price -- then a bubble exists.”
Joseph Stigliz (1990). “Symposium on Bubbles,” Journal
of Economic Perspectives 4(2), Spring, pp. 13-18.
Determinants of housing prices:
Demand and supply factors or “fundamentals”:
Interest rates, financing conditions, household income,
taxes, operating costs, population growth and density,
demographics, consumer and builder confidence,
new construction and development costs, regulations,
land supply, geographic constraints, information and
search costs, investment demand, expected price
appreciation, opportunity cost, quality of life,
non-monetary value of homeownership.
Low readings for flipping/speculation indicator
Holding Period of Residential Sales: Metro Vancouver
Per cent of sales
50
3-months
6-months
1-year
40
30
20
10
0
Jan-79 Jan-84 Jan-89 Jan-94 Jan-99 Jan-04 Jan-09 Jan-14
Source: Landcor Data Corp.
Latest: Mar-15
Some observations:
 Speculation in housing increases in robust market
conditions with rising housing sales and prices
 Speculation drops when housing cycle turns into
recession phase
 Most speculative market was 1979-81 followed by 1987-
89 and 2005-08
 Lower speculative activity peaks over time
 Current flipping activity is very low
Some takeaways on myth #1:
 Price-to-income (or rent) ratio incorrect valuation metric;
widely misused
 Long term average of ratio is meaningless benchmark;
market forces do not cause reversion to mean especially
in markets with rising long-term price trend
 Housing valuation is complex exercise requiring more
detailed analysis and data than readily available
 Housing prices determined by many demand/supply
fundamentals; supply factors almost always overlooked
 Land supply constraints too important to be ignored
Housing Myth #2:
 Vancouver is among the least affordable markets
in the world
Some commentary on Vancouver housing prices:
“As is well known, Vancouver is the second-least
affordable housing market in the world after Hong
Kong..., Vancouver Sun, Sept. 4, 2014
“Vancouver leads the pack at a whopping 9.5 — the
second highest in the world behind only Hong Kong…”
The Ottawa Citizen, Sept. 11, 2014
“Vancouver ranks second only to Hong Kong in having
the least affordable housing, according to Demographia's
10th annual survey of 360 housing markets in nine
Western countries.” CBC News Jan. 21, 2014
Vancouver is the second least affordable behind
Hong Kong
Source: 11th Annual Demographia International Housing
Affordability Survey: 2015 pp.33
Some data issues with Demographia report:
 Only nine countries are covered; there are more than
190 countries in the world.
 Some notable countries not covered are France,
Germany, Holland, Switzerland, and the Nordics.
 Not all metropolitan areas in some countries are
covered, for example Osaka-Kobe-Kyoto in Japan.
 Data limitations prevent Demographia from including
more countries and metropolitan areas.
 Nonetheless, some commentators jump to the
conclusion this report covers the entire world
Some data issues with Demographia report:
 Housing is vastly different in form (dwelling type and
size) across countries and metropolitan areas.
 The median price used in the report is an aggregate of
all housing forms but some metropolitan areas are
mostly high density housing (apartments) and others are
mostly lower density (single family homes with land)
resulting in an apples-to-oranges comparison.
 High density metro areas are Hong Kong, Singapore,
Tokyo, and other metro areas not included in the
report. In Metro Vancouver, most housing is single
family dwellings on freehold lots and only 40% are
apartments.

Some data issues with Demographia report:
 Most of the metro areas in North America covered in the
report are largely low density housing.
 The average size of a home in high density areas is
considerably less than in low density areas such as
Metro Vancouver. For example, there are reports of
apartments less than 100 sq. ft. in Hong Kong.
 The median price does not correct or adjust for these
housing form and size differences. A more accurate
measure would be median price per area but that
statistic is very difficult to compile.
Some data issues with Demographia report:
 The report uses the MLS average sale price for its
Canadian metro areas and converts these averages to
medians by a historic conversion factor.
 It estimates Vancouver’s median price at $704,800 in its
latest report. However, the REBGV median residential
sales price is actually $570,400. Demographia’s
conversion factor is incorrect.
 Further, the growing divergence between the average
and median sales prices in Vancouver makes this
analysis even more dubious.
Some data issues with Demographia report:
 Another problem is the coverage of metropolitan
areas. Vancouver’s price is based on the Real Estate
Board of Greater Vancouver MLS data but this board
area is smaller than the official Vancouver metropolitan
area -- Surrey and Langley are excluded, as well as
North Delta and White Rock.
 This overstates Metro Vancouver’s actual median price
and distorts comparisons with other metro markets.
 The report includes the Fraser Valley (the Fraser Valley
Real Estate Board area), which includes Surrey, Langley,
North Delta, Abbotsford and Mission.
Some data issues with Demographia report:
 Metropolitan areas in the nine countries reported are not
equal. For example, the New York metropolitan area
covers some 35,000 sq. km. and a large area of lower
density housing while Metro Vancouver is 2,900 sq. km.
with critical geographic constraints on urban
development.
 Comparing prices in the Central Business District or 20
km or 40 km from the CBD of each metropolitan area
would be more revealing and accurate.
 Ideally, land prices should be used because it reduces
housing form differences.
The author’s cautionary note:
 “House Characteristics: The indexes and data on which
the Survey is based reflect the overwhelming majority of
existing housing in the markets. At the same time, there
are differences in house types, housing characteristics
and lot size between the geographies covered. The
Demographia International Housing Affordability Survey
does not adjust the Median Multiples to reflect these
differences. For example, the average size of housing,
particularly new housing, is abnormally small by New
World standards, in the United Kingdom and Hong
Kong.”
Prices of apartments in city centre
Rank
1
2
3
4
5
6
7
8
47
65
City
Hong Kong
Kowloon
Singapore
Geneva
London
Lausanne
Zurich
Taipei
Vancouver
Toronto
Price/ sq.
metre USD
$27,113
17,129
16,267
15,970
14,927
13,368
12,291
11,203
6,453
5,293
Source: Numbeo. http://www.numbeo.com As of Apr. 2015
Price-income ratios and rankings
Rank
4
9
19
33
64
180
187
203
231
262
City
Hong Kong
Kowloon
Taipei
Singapore
London
Lausanne
Geneva
Vancouver
Toronto
Zurich
Price/income
36.83
31.10
25.79
21.76
16.56
10.38
10.13
9.61
9.04
8.07
Source: Numbeo. http://www.numbeo.com. As of Apr. 2015
Some takeaways:
 Vancouver housing is not the second least affordable in
the world
 The Demographia report does not make that claim,
rather some commentators jump to that conclusion
 There are many more expensive cities than Vancouver
and Toronto
 Making international house price and affordability
comparisons is loaded with data issues and availability
 That Vancouver and Toronto housing is expensive or that
ownership is unaffordable for many is not in question.
Housing Misconception:
 High prices and poor affordability cause price
collapse
Housing prices drop during economic recessions
Change in Housing Prices: Vancouver and Toronto
Percentage change at annual rate
120
100
80
60
Vancouver
Toronto
40
20
0
-20
-40
Jan-77
Economic recession
Jul-84
Jan-92
Jul-99
Jan-07
Jul-14
Source: CREA, Landcor Data Corp, Central 1 Credit Union. Latest: Mar-2015
Some observations:
 Economic recessions cause housing recessions. Job
and income losses, drop in confidence, and migration
outflow causes declining housing sales
 Economic recessions in Canada usually occur due to
external economic or geopolitical shock events.
 Economic and housing recessions are temporary.
 Housing supply adjustments occur leading to price
stability.
Some observations:
 The main risk of a housing crash is from an economic
recession – a sudden drop in demand.
 Housing crash predictions based on high price-income
ratios reverting to mean are made on inadequate
analytical and empirical grounds.
 High housing prices alone do not cause housing
recessions.
 Housing prices will decline during the next economic
recession.
Key takeaways:
 High housing prices and poor affordability result in
market adjustments to demand and supply.
 Smaller units are constructed, rental demand increases,
and some household ‘double up’.
 In Metro Vancouver, about 65% of households are
homeowners and 35% are renters.
 In the next 30 years, this changes to less than 60%
owners, and more than 40% renters.
Thank you