Last fall Paris based bank, Societe Generale, produced a comprehensive report on the impact of changing global demographics on consumer trends.  The report points to the rapid development of property markets in much of the western world since the 1990's as a by-product of maturing baby boomers shifting into an asset-accumulation-phase of their consumer life.  Although the drivers of the Canadian housing boom are best explained by the coalescence of many factors, demographics shifts are often an underappreciated and often misunderstood contributor.

To better understand current consumer age demographics in Canada, Chart D1 below dissects Canada’s current population by age and consumer behaviour groups.  As displayed in the chart a large bulbous mass of “experienced workers” will enter retirement in upcoming years - a fact that should come as no surprise and one that is frequently cited.  These retirees are generally asset rich but with decreasing ability to generate revenue and with reduced savings potential.  But most relevant to real estate, retirees finance consumption in large part through depletion of both: liquid-savings and illiquid equity in hard assets (eg. real estate).

Chart D1: - new



An interesting observation by Societe Generale (Soc Gen) is that most housing bubbles in industrialized nations appear when the share of the pre-retiree population (i.e. aged 45-65) experiences rapid growth. These two events, housing price appreciation and growth in the 45-65 year old population, are - as the report states, "no coincidence".  In other words, the run up in housing prices over this period of shifting demographics is largely explained by the same group's changes in consumer demand.

Applying Soc Gen's observation to present day Canada, we uncovered a significant break in trend over the last two decades.  The inital trend, as shown in Chart D2 below, involved the 45-to-65 year old cohort which remained a relatively constant proportion of the total Canadian population at just below 20%. The trend continued for two decades, the 1970’s and 1980’s.  Yet over the following two decades, the 1990's and 2000's, this same cohort swelled to contain 10% more of the overall Canadian population.   In absolute numbers, there are approximately 4.3 million more Canadians in the 45-65 age group now than there were in 1990.  The size of the shift in demographics is almost equivalent to the current population of Canada's third largest province, British Columbia. What should be taken from this is that the shift in proportional population sizes had a tremendous impact in aggregate Canadian consumer behaviour.

Chart D2: - new



The other interesting observation from the chart above is the decreasing supply of "young dependents", or youth under the age of 20, that is now at generational lows.  This cohort in past generations was plentiful enough to not only replenish older workers exiting the Canadian labour force, but also fueled growth of the labour force itself.  What is the impact of this shift? Simply put, the potential labour force in Canada is forecast to shrink. Barring changes to immigration, Canada is forecast to see an increasing "dependency ratio", or the number of child dependents and elderly in relation to the potential labour force (Chart D3 below).  These dependents will effectively act as a social tax resulting from a shrinking labour force carrying the social costs of a rapidly aging population.

To fully appreciate Canadian historic and projected demographic shifts and their impact on real estate pricing, one should pay attention to Chart D3 below.  The chart identifies that the economic boom witnessed in Canada over the last decade was not only a function of global commodity demand and increasing leverage, but was also assisted by the tail end of a massive demographics boom.  This demographics boom began in 1965 and lasted until 2010.  The by-product of the boom was rapid growing demand for consumer products, consumer services, real estate and anything else coveted or required by a young growing work force powered by expanding credit. Assuming United Nations' population forecasts for Canada hold, Canada's shifting demographics will likely result in a headwind for some consumer-sensitive components of the domestic economy. Components of the economy that once serviced the needs of a younger population but are out of favour by the elderly will come under pressure.   It is reasonable to assume that future Canadian consumer behaviour will be markedly different than what was witnessed over the last half century.  Please note, the reliability of even basic demographic forecasts begin to severely erode after approximately 20 years, however, for illustrative purposes we have included the UN's projections to 2100 in the chart below.

Chart D3: - new





Real Estate Review - The last six months

Since our last chartbook update in July 2012, the Canadian housing market has been anything but uneventful.  The headline news of the period was the change to mortgage rules by the Finance Minster, Jim Flaherty, that were put into effect in early July with the goal of curbing aggressive expansion of Canadian mortgage debt (read more).
 
In the same month that new mortgage rules were put into place, Standard & Poor’s Rating Services revised its credit outlook from stable to negative on several Canadian Banks due to exposure to consumer debt and unsustainable home prices (read more). The change in S&P’s credit outlook was in addition to a later warning by credit rating agency Moody’s, indicating that it too was reviewing the credit rating of six Canadian banks and downgrades could follow (read more).  Additionally in October the International Monetary Fund (IMF) cut Canada's economic growth forecasts also citing overvalued Canadian housing market and heightened consumer debt levels (read more).

Even esteemed Yale economist and US housing expert Dr. Robert Shiller weighed in on Canadian housing in a CBC interview in September in which he stated, "I worry that what is happening in Canada is kind of a slow-motion version of what happened in the US" (read more).

At the data level, Credit reporting agency TransUnion reported that Canadian debt loads grew at the fastest pace in two years during the Summer of 2012, despite numerous warnings by Mark Carney and Jim Flaherty to Canadians on the risks of excessive leverage (read more).  Statistics Canada also released data revisions indicating that Canadian debt had risen to 163% of disposable income, placing Canadian ahead of indebted Americans and British consumers on credit consumption (read more). 

Added to all of this activity was yet another standard-issue warning by Mark Carney, Governor of the Bank of Canada, on the dangers of excessive household debt (read more).  Unfortunately, Governor Carney's warning came only weeks before announcing he was leaving the Bank of Canada for the equivalent seat at the Bank of England, a country that has already experienced its housing correction. 

Our gauge of consumer real-estate sentiment, powered by Google search volume for bearish real estate terms, spiked in mid 2012 as the aforementioned events began to unfold (Table 1).

Significant price weakness also occurred. In Vancouver, (Chart 11), quarter over quarter home prices have experienced the largest drop since the 2008 financial crisis and the second largest drop in a decade.  Average single family detached home prices, which can be skewed by non-representative sales, have dropped almost 13% between April and October.  Meanwhile price appreciation for Vancouver town houses and condos is non-existent with current prices nearing early 2008 levels and trending lower (monthly averages courtesy of Brian Ripley).  Yet, Vancouver still remains 38% overvalued when compared to its historical valuation ratio (Table 2).  Edmonton, Calgary, Toronto, and Montreal appear similarly overvalued. 

Policy Analysis: Errors and Heroes

The recent decrease in home sales has been incorrectly blamed on the July change in government insured mortgages.  Under most scenarios, the reduction in max amortization had the equivalent impact on monthly mortgage payments of roughly a 1% increase in mortgage lending rates.  Thus the weakness in housing that followed sent a clear signal that the real estate market was dependent on low rates and expanding credit to fuel further price appreciation.  The tightening of amortization periods served more as a real-time stress test for the Canadian real estate market than an attack on the Canadian consumers.  Unfortunately in this stress test Canadian real-estate failed to display any resilience.  

The maximum insured mortgage amortization period that took effect in July is now 25 years. It was reduced from 30 years which itself was set after an earlier reduction from 35 years.  Yet an even earlier reduction in insured mortgage max amortization periods from 40 years to 35 years was put into effect in 2008.  The recent tightening of max amortization, however, was not anything new.  Instead, the changes only served to reverse a 2006 decision by the then governing Federal Conservative party to raise the maximum amortization period from the then 25 year max (what it is again now) to 40 years.  This one decision may be remembered in Canadian economic history books as an extremely grievous policy error.

Prior to the 2006 loosening of government insured mortgages; the Canadian real estate sector had already experienced a half decade of surging prices.  The National Bank – Teranet home price index for the aggregate Canadian housing market indicated a 50% rise in home prices between January 2000 and December 2005 (Chart 27).  Thus the policy change of 2006 was counter to basic Keynesian economics which advocates the use of fiscal and monetary policy for smoothing of business cycles. This is achieved by stimulating slow growth environments and enacting policies to cool excessively high growth economic environments.  The 2006 policy change, however, served a contrary goal of stimulating an already surging market.  What followed was a 26% increase in outstanding mortgage credit between mid 2006 and mid 2008 (Chart 2).  Meanwhile, Canadian nominal GDP grew by half as much, only 13% (Chart 5) and weekly Canadian Wages grew at slightly more than one-quarter at only 7% (Chart 6).

In a recent statement the Canadian Finance Minister, Jim Flaherty, is now attempting to take credit for his recent actions, which in truth, only served to reverse his critical 2006 policy error.  Minister Flaherty exact statement was, "The housing market has softened somewhat in part because of steps that I've taken and I'm happy about that," (read more).   In truth, the only policy maker in Canada that should be held in high regard is former Bank of Canada Governor, Dr. David Dodge, who in June of 2006 angrily criticised the Canada Mortgage and Housing Corp for utilizing new mortgage products which would lead to housing price inflation and reduce affordability (read more).  It is now apparent that Dr. Dodge was correct.

The fall of 2012 also witnessed the start of a Canadian real estate market slowdown as observed through rising unsold-home inventory levels, and falling sales volumes.  Even in Canada’s previous hotbed of real estate activity, Vancouver, sales volumes slowed significantly and inventory builds followed.  The Real Estate Board of Greater Vancouver reported that November 2012 sales were 28.6% lower than they were the same month last year (read more).   The Toronto Real Estate Board reported similar findings with November mid month sales down 17.5% with the same period of 2011 (read more).

The drastic slowdown in sales volume in some Canadian markets has important impacts on future pricing. This is because a slowdown in home sales can create a build in unsold housing inventory as we are seeing in various regional markets.  As with any consumer good, in order to liquidate excess inventory back to a long term average, the price of the goods must be reduced. Essentially, homes have to be put on bargain pricing to entice buying and reduce inventory.
 

Homeowners are naturally reluctant to lower prices and will first attempt to remove listing and re-list in a more favorable environment.  However, home prices are determined not by idle inventory but the price that units are exchanged at.  During periods of decreased sales volume, the few sales that occur will set the prices for the entire market.  This is what took place in the US real estate correction over the last half decade.

Table V1 below displays the S&P Case Shiller home price index for seasonal “sales pair” data, an approximation for aggregate home sales in major US housing markets.  The table tracks the relative proportion of sales per year and per season to the same period's ten year moving average.  What we glean from this is that the surge in US home prices leading up to the 2006 housing crisis was on the back of increasing sales volume. Alternatively, the severe drops in US home prices occurred during periods of shrinking sales volume.  Effectively, US home prices dropped on a housing liquidity drought in which home prices for the entire market were determined by a shrinking number of transactions  at lower and lower prices.  The US home sellers who refused to budge on price themselves, had their homes forcibly re-valued by the market through lower priced comparable sales.  

Click Here to view a larger version of this table.

An update to this Chartbook will be published in April/May 2013.

Canadian Real Estate bulls have continued to cite the fact that real estate valuations have appeared "expensive" for years, yet, the momentum has continued to take prices higher.  Real estate bears, on the other hand, claim that home prices have been so stretched from fundamental valuations that past price momentum is irrelevant.  As with all asset classes, a change to investor sentiment regardless of the catalyst that triggers the change (eg. rising interest rates, government policy, extreme valuations, etc.) will dictate future real estate returns.  

To attempt to monitor real estate sentiment analytically, we examine the proportion of Google searches for the combined terms "housing bubble", summarized by the originating city of the searches.  The table below displays the top six cities globally in which the term "housing bubble"  searched in each year from 2005 to 2012.  The highest proportion of city-sourced searches for "housing bubble" arise from two Canadian cities, Vancouver and Toronto.  The progression of the trend is ominous as California cities dominated much of the top searches until 2009, after which Canadian and Australian cities began to emerge.  Absolute search volume for all of Canada is also plotted.  Note, the spike in search volume in mid 2012 corresponding with a slow-down in housing sales volume and mortgage lending rule changes.

The data below reflects recent updates to Google-Trend's data.

Despite falling interest rates, Canadian consumer credit growth has slowed to the lowest levels in more than 12 years.  This observation is despite the fact that real bond yields, or inflation adjusted yields, have dropped significantly.  The axis on the right in red tracks 3 to 5 year real Canadian government bond yields which are now below zero.  In other words, interest income on these bonds are no longer sufficient to overcome lost purchasing power from the effects of inflation.  Should consumer and mortgage credit begin to contract then this will serve as a major headwind to future real estate appreciation.  

Canadian household credit, both consumer credit and residential mortgage credit, has increased sharply over the last decade.  Total credit as a percentage of nominal GDP increased from under 60% during the third quarter of 2001, to the current levels of 90%.  The surge in Canadian household debt is predominantly comprised of residential mortgage credit as can be observed from the chart below.  Canadian consumer debt accumulation appears to have been exhausted and is currently in a stall state of approximately 90% of GDP.  Lack of debt accumulation is a headwind for further real estate price appreciation.  

In all major Canadian housing markets housing capacity growth has exceeded population increases between 2001 and Sept 2012.  Calgary, Edmonton, Ottawa, Montreal and Halifax are what we would consider to be "severely overbuilt" with excess housing capacity of 50% or more than population growth over the same period.  Vancouver, Toronto, and Winnipeg, are "overbuilt" with excess housing capacity 20% more than population growth over the examined period. 

Housing capacity is defined as the number of individuals that can be reasonably housed in new housing units, whether or not a new housing unit sits unoccupied, under-occupied, or over-occupied.  Assumptions made may be more appropriate for some markets over others.  

In theory, residential real estate prices should equal the discounted sum of future rental income.  As a result, we have attempted to estimate fair values for residential real estate in major cities by comparing actual prices to theoretical discounted prices (valuation ratio).  In theory, this ratio should equal one and deviations from this value should regress back to the value one over time.  Note, discounted cash flow calculations are highly volatile and dependent on underlying model assumptions.  However based off of this methodology, Canadian real estate appears extremely expensive in most major markets.  Canadian real estate only appears somewhat reasonably priced if the assumption that current emergency low interest rates continue indefinitely into the future.  Any increase in interest rates to even pre recession levels (which were also historically low) causes Canadian real estate as a whole to appear grossly overvalued.

Table 2 below attempts to quantify the degree of price correction necessary to return the valuation ratio  in these five real estate markets back to the historical average valuation ratio.  

Using the data from Chart 5 above, the following table attempts to quantify the degree of price correction necessary to return the valuation ratio in these five real estate markets back to the historical average valuation ratio.  The price corrections necessary to return the valuation ratio to the historical moving average range from -39% in Vancouver, to -31% in Edmonton.  

Canadian wage growth versus home price appreciation from 2001 to 2012 is reported below.  Average weekly wage growth per home province of each city is reported.  Also, only wages of full time workers between the ages of 25 and 54 were examined in an attempt to capture changes to the buying power potential of first-time-homebuyers.  In all ten markets examined, home price appreciation far surpassed average weekly wage growth.  The markets with the largest difference between wage growth and home price appreciation in descending order are:   Winnipeg, Vancouver, Montreal, Edmonton, Calgary, Ottawa, Toronto, Halifax

The Canadian misery index (inflation + unemployment rate) had been climbing since hitting a low at the end of the first quarter 2008.  It has since peaked in 2012 as a result of lower inflationary pressure. Calgary is approaching multi year lows in misery.  The three largest Canadian cities have misery levels above the lows reached prior to the 2008 financial crisis. 

Long term real (inflation adjusted) annual home price returns have exceeded 3% in Vancouver and Victoria BC, while exceeding 1.5% in most other large Canadian cities.  Edmonton is the only exception with a compounded annual house price appreciation of 0.64% over the examined period.  To put this into perspective, numerous examinations of long term real US home price appreciation indicate that they have only slightly exceeded inflation at an approximate annual compounded rate of 0.5% per year.  

Canadian home prices are currently not in line with historic multiples of residential rental prices.  Most extended from historical norms are Vancouver, Montreal, and Toronto.  While Edmonton and Calgary, are elevated from historic averages but below previous witnessed highs.  

Displayed in the chart below are Canadian home prices as a ratio of the TSX index (Canadian stock market) from 1977.  Seven cities are included: Vancouver, Victoria, Calgary, Edmonton, Regina, Toronto, and Montreal.  Over the long term, home prices in Canada have lagged price appreciation of stocks.  Note, the stock index below is a "price index" and therefore, excludes payment of dividends.  

Displayed in the chart below are Canadian home prices as a ratio of the TSX index (Canadian stock market) from 1998.  Nine cities are included: Vancouver, Victoria, Calgary, Edmonton, Regina, Ottawa, Toronto, Montreal, and Halifax.  Over the medium term, home prices in Canada have outperformed price appreciation of stocks.  Note, the spike on the charts observed at March 2009 represent the stock market bottom during the financial crisis.  

For comparison purposes the following two charts (Chart 20 and Chart 21) have also been included which display US home prices as a multiple of the S&P 500 (US stock market).  The chart immediately below displays US home prices as a ratio of the S&P 500 index (US stock market) from 1987 onward.  Fourteen US cities are included in the chart below as well as a composite index of ten major US Cities.  Over the medium term, home prices in Canada have outperformed price appreciation of stocks.  Note, the spike on the charts observed at March 2009 represent the stock market bottom during the financial crisis.