CANADIAN HOUSING: FIRST IN, FIRST OUT,
BUT WHERE TO FROM HERE? (from TD Economics.Com)
Of late, the Canadian housing market has been the focus of a lot of attention
from the media and analysts alike, and for good reason. TD Economics has itself
commented on these developments in two recent pieces1 , providing its residential
real estate forecast and highlighting the potential monetary policy implications of
various scenarios. The objective of the current note is twofold. Firstly, we update
the outlook to incorporate data made available since our last report. Secondly, we
expand on specifi c concerns that have arisen in the current housing market context.
Through the ringer – as sharp as ever
Existing home sales and prices, as provided by the Canadian Real Estate Association
(CREA), went through a sharp downturn last year, falling by 40% and
12% respectively from their peak of late 2007. Just as quickly and sharply, a phenomenal
rebound kicked off early this year and was still going strong early in the
fourth quarter. From their trough sales had surged by 74% as of October, while the
average price was 20% higher.
In this extremely sharp two-year cycle, the housing market has undeniably
held up to its ‘fi rst-in, fi rst out’ (FIFO) historical billing. The downturn in existing
home sales and prices (Q1/2008) preceded the start of the technical recession
(dated Q4/2008) by three quarters. On the fl ip side, the strong recovery in existing
home sales and prices that started in earnest in Q1/2009 led the end of the technical
recession (dated Q3/2009) and the start of the overall economic recovery by
at least two quarters.
Furthermore, the Canadian resale housing market downturn and recovery was
as V-shaped as can be. A quick glance at the chart at the top of the next page shows
sales and the average price could make a calligrapher green with envy. The length of
each leg of this cycle was also nearly symmetric with the downturn lasting roughly
all of 2008 and the ensuing recovery spanning all of this year.
While in the thick of a recession, the strongest countervailing force that set the
stage for the mother of all rebounds, apart from lower prices, was lower interest
rates. Recall that the Bank of Canada began easing its monetary policy back in late
2007, when it was becoming clear that the U.S. economy was tilting into recession
and would surely drag Canada along with it. By the time the recession offi cially hit
in Canada a full year later, the overnight rate had already been slashed from 4.50%
to 2.25%, more than halfway en route to its all-time low of 0.25% by April 2009.
All said, the housing market has gone beyond retracing its steps and fully
recovering from the end of 2007 – which had marked the peak of a half-decade
long boom, concentrated in Western Canada. As of October 2009, national sales
were running at a blistering 550K annual pace and the average price was $340K.
Seasonally-adjusted monthly sales hit all-time record
volumes for 3 of the last 4 months, and are on track to continue
this record-setting pace over the next few months. As
of October, both sales and the average price stood 5% higher
than their respective 2007 peak. Extrapolating this trend
echoes Buzz Lightyear’s mantra “to infi nity, and beyond!”.
Back here on earth, however, this latest housing cycle
raises a number of concerns. For one, was the whole twoyear
cycle just a blip? A second, related question would be: is
the recovery sustainable or was it all too much, too fast in the
midst of a recession and early stages of recovery? Digging
into the pace and magnitude of the rebound, more technical
questions arise, such as: how much of the rebound was simply
the unleashing of pent-up demand? Alternatively, how
many of the current sales are simply being brought forward
on the expectation that interest rates must eventually rise,
in effect stealing from future demand? Last but not least, is
a bubble brewing in Canadian housing?
Just a blip?
The accompanying chart showing sales and the average
price suggests the recent cycle was just a blip. Affordability,
as measured by typical mortgage payment as a percentage
of average household market income, is also reverting back
to where it stood before the downturn. After improving
from 32.4% to 26.2% in 2008, it was back up to 29.5% in
Q3/2009. We forecast this measure will have climbed right
back to 32.6% by Q4/2011, thus erasing all of the improvement
in affordability seen during the downturn.
The ‘blip’ story goes something like this. The downturn
in 2008 was that of a housing market being just an innocent
by-stander getting unfairly sideswiped by fears of a U.S.-
style downturn and rock bottom consumer confi dence in the
midst of an extreme fi nancial crisis. As these fears abated
and the worst of the fi nancial crisis passed, the doom and
gloom headlines became more nuanced and some modicum
of confi dence was restored, whereas little existed before. As
a consequence, the housing market navigated rough waters
and made it through the storm relatively unscathed.
This narrative is not wrong per se, but its implicit conclusion
is that all is back to ‘normal’, whatever that may be. It
also suggests that it will be smooth sailing from now on. In
other words, the volatility in sales and prices has shaken out,
and the expectation from here onward is a steady uptrend.
As a result, we see two problems with calling this cycle
a blip. Both arise from near-sightedness. First, it fails to
think about where the market stood pre-downturn. Second,
it neglects the fact that the current uptrend is too steep and
that the resulting erosion in affordability will come back to
bite into future demand.
To expand, the fi rst problem is that the ‘blip’ notion
fails to take a longer-term perspective on home values. If
the Canadian housing boom (roughly 2002-2007) resulted
in modestly overvalued residential real estate, which we
estimate was roughly 10% 2 on a national scale at the 2007
peak, the downturn had just about set things right from a
value perspective, with a peak-to-tough price adjustment of
-12% 3 . For a number of reasons, we never forecasted a
U.S.-style crash in Canadian housing. On the other hand, the
adjustment that took place in 2008 looked warranted from
a fundamental value perspective. The resulting improvement
in affordability that came from more modest prices
was encouraging and sustainable from a macro-fi nancial
perspective. From their trough, the most sustainable path
for Canadian home prices would have been a gradual and
modest uptrend aligned with nominal income growth. But
now that home values are already past their previous peak in
such short order, we estimate that the typical home remains
overvalued by 12% at the national level. Unfortunately,
sheer momentum suggests that this overvaluation is likely
to increase over the course of the next few quarters, peaking
at 13-15% in H1/2010.
The misalignment of home prices with their fundamental
drivers, such as demographics and income, cannot last. That
much is known. What is less clear is the exact timing of
when and precise channel by which the two will eventually
realign. Because a necessary realignment has been erased
so quickly without support from income growth, another
adjustment must take place – although it could take many
forms. As of our writing this note, early signs of market
cooling are emerging and our analysis still suggests the
most likely outcome is a soft landing and relative stagnation
of home values in real-terms along with a resumption of
stronger income growth over the 2011-13 time frame. Turn
to our forecast section for the specifi c profi le projected over
the next couple of years.
The second problem with the ‘blip’ characterization is
that fails to look forward to the eventual resetting of interest
rates – what happens when the sturdy trampoline of rockbottom
policy interest rate that continues to fuel the sharp
market rebound is taken away? Changes in affordability
that rest solely on lower interest rates are inherently cyclical
in nature, as opposed to those that arise from household
income or homes prices.
Call this housing cycle a blip if you like. But we feel that
is misleading, especially because of what this suggests for
the future. While the market looks remarkably unperturbed
from start to end of this sharp cycle, existing home sales and
prices cannot sustainably stay on their current path. (See accompanying
chart for 3-month trend in M/M annualized %
change price). Markets are currently very tight and favour
sellers, as evidenced by multiple competing offers and bidding
wars, but we expect them to rebalance over the course
of 2010. As the central bank begins to hint at a tightening
monetary policy cycle in the second half of next year, sales
could well see a last gasp of strenght. Moreover, by that time,
the availability of units on the supply side should provide
a relief valve helping to cool price growth. And, by 2011,
while the overall economy will have improved signifi cantly,
housing markets will be losing momentum
Repaying the past, stealing from the future
On the issue of pent-up demand, we had calculated that,
on a nationwide basis, at most 53,000 existing home sales
that would normally have occurred in Q4/2008 and Q1/2009
did not occur because of the crisis of confi dence resulting
from the fi nancial market turmoil. This fi gure is established
on the basis of a continuation of the pre-recession downtrend
in sales, which actual sales undershot signifi cantly. Since
Q2/2009, however, sales have overshot that trend by a wide
margin. In our previous piece, we estimated that 50-60% of
that pent-up had been released as of August. With two more
months of data now available, we calculate that 75-100%
of this pent-up demand would have been absorbed by October.
Sales have been tracking our near-term expectations
well and we continue to judge that any remaining pent-up
demand will have likely been exhausted by November. At
the very latest, by year-end this source of demand will have
completely dried up.
The full absorption of pent-up demand by itself should
help to slow overall sales in the fi rst half of 2010 compared
to their recent pace, which has already begun to cool on a
3-month trend basis. (See accompanying chart). Over the
course of Q2/2009 and Q3/2009, up to one in fi ve sales
(monthly seasonally-adjusted units) could reasonably have
been attributed to those that had previously been delayed
(pent-up) because of sheer uncertainty.
While an important factor, this is clearly not the single
or most important factor fueling overall demand. Demand
has mostly been supported by attractive fi nancing rates
which have more than offset the headwind created by weak
labour markets. As this is not expected to change much in
the near-term, we do not anticipate sales to simply drop by
a fi fth come January – which is what would happen if other
supporting factors were lacking. Nonetheless, it serves as a
useful gauge of underlying drivers separate from displaced
demand coming back online. In our view, any sales observed
from January 2010 onward will originate not from past
displacements of demand, but from traditional real estate
drivers. This should enable us to get a much better reading
of the underlying strength of demand after transient factors
have washed out.
A more diffi cult issue relates to how much of the current
demand is simply being brought forward, i.e purchases
in recent months that advanced sales to take advantage of
low rates, which raises the risk of a dip in ensuing quarters.
Because the pool of potential buyers is not fi xed and itself
depends on affordability, it is not possible to satisfactorily
address this issue in a precise quantitative fashion manner
with the current data available. There is little doubt as to
the direction of this effect, however. The prevailing ‘now
or never’ mentality will weigh on future demand.
Too much, too fast?
The speed and magnitude of recovery has been a surprise
to all. After all, it occurred in the midst of a recession during
which unemployment rose signifi cantly. Incomes also took
a hit, particularly those tied to slumping commodity prices.
The combination of declining home prices and lower interest
rates dramatically improved home affordability over the
course of 2008. And while affordability has no longer been
improving since Q2/2009, the impact of past improvements
in affordability is still rippling through resale markets and
helping to spur sales.
On the home price front, any answer to the sustainability
question must distinguish between current levels and current
momentum. While current price levels are above what
we estimate to be long run fundamental values, they do
not appear so dramatically out of line as to warrant a sharp
correction in the near-term. Such corrections are typically
triggered by a macro-fi nancial event such as we saw when
yields spiked in 1994 or during the fi nancial turmoil unleashed
in the fall of 2008. This risk of corrections always
lurks, but a stabilization of prices around current levels
could be sustained, as affordability would remain decent.
As for price momentum, it is more clearly unsustainable.
On a 3-month average M/M annualized percent change,
average home price growth was 22% in October, but has
been declining since peaking at over 40% in July. We expect
double-digit growth by this measure to wash out after
Q1/2010. Recall that every price increase that is not matched
by a commensurate income gain increases the overvaluation
gap. Second, more supply should come online in the fi rst half
of 2010 in the form of new home and condo completions.
While the number of units under construction remains much
lower than a year prior in most urban markets, they near an
all-time high in the Toronto area (see accompanying chart),
which will provide a supply relief valve in Canada’s single
largest market. In reaction to the recent price gains, we also
expect a positive supply response on the existing home front,on this front as this uptrend has yet to materialize. Lastly,
sometime in the second half of next year it will become
evident that interest rates must rise, which is expected to
dampen sales considerably in 2011-12 when compared to
sales expected for 2010.
Bottom line – housing outlook 2010-13
All said, it looks to us as if the rebound was a tad overdone,
but it is not so much the current level of prices which
raises concerns. What raises eyebrows is where the current
market momentum will bring prices next year. The current
market tightness, as measured by the sales-to-listings ratio
(see accompanying chart), while expected to ease gradually
over the course of 2010, will not turn on a dime. As a
consequence, it will be supportive of price growth in 2010
that is stronger than fundamentals can support over the long
haul. After climbing by an estimated 4-5% on an annual basis
this year, the average existing home price is expected to gain
another 9-10% in 2010 as sales climb to 475K.
But the current momentum is not expected to last beyond
the next 6-10 months. Were it to continue into 2011,
there would be more credence to the view that a bubble has
formed. But the brakes are currently being applied in the
background, which should prevent a bubble from forming
between now and then. Measured in terms of affordability,
Q3/2009 marked the third worst deterioration on record –
which dates back to Q1/1988. Previous historical episodes
(in 1989 an 1994) caution that the market could stall in
upcoming quarters. While interest rates may rise, they are
unlikely to spike as they did back then, which provides the
market with better shock absorbers than in the past. Nonetheless,
as supporting factors wane and affordability erosions
weaken sales by over 10% in 2011, prices will struggle to
keep up with CPI infl ation.
As interest rates continue to normalize to higher levels
in 2012 and the economic backdrop continues to improve
2012-13, sales are expected to climb only modestly during
those two years to reach 450K by 2013. A larger supply in
the form of new and existing units should weigh down on
nominal price growth to the point were we expect real prices
(adjusted for infl ation) to stagnate as incomes are fi nally allowed
to catch up to home values. Assuming annual nominal
income growth of 4-5%, home overvaluation would wind
down to 4-8% by Q4/2011 and would essentially vanish by
Q4/2012 under this forecast profi le.
In closing, we note that the most important downside risk
to our near-term forecast is not that the market cools more
than we anticipate. While this risk certainly exists, it would
not cause signifi cant market disruptions, and it would ensure
that affordability does not continue to erode at the current
pace. The risk is rather that the market remains as hot as it
currently is for too long, eventually running head-on into
monetary policy tightening (and longer term bond yields
rising). There is more than adequate time for the housing
market to cool before then, but history suggests that if it fails
to do so, the ensuing adjustment would be a rude awakening.
Longer term, Canadian households also need to ease
debt growth to bolster net worth when asset price growth
moderates. Debt-servicing costs will undoubtedly rise over
the next few years. While most households can handle this
rebalancing act, those already overstretched or getting into
homeownership on the margins of affordability would do
well to plan ahead by building up equity and saving through
other means. We will also be examining these issues in detail
in a forthcoming paper.
although it must be said that we make a prudent projection
Canadian Housing: First In, First Out, But Where to From Here?
