Housing prices: Accelerated boom cycle not anticipated
Aside from a few hotbeds like Toronto and Calgary, Canada’s housing boom effectively ended in 2008.
Since then, prices have largely kept pace with incomes, consistent with a soft landing.
This note examines the economic impact of the earlier boom, with a view to sizing the possible damping effect in the event of a correction.
After a prolonged period of weakness in the 1990s, home prices rose five percentage points faster than personal income from 2002 to 2007 on average.
The 72 per cent increase in prices over six years lifted household wealth, confidence and borrowing ability, ergo supporting the economy.
Using BMO’s quarterly GDP model, if no boom had occurred and prices simply rose in line with personal income, the economy would have grown 0.22 percentage points slower per year on average.
The supportive impact contrasts with that in the previous decade, when a prolonged period of price stagnation shaved 0.15 percentage points from annual growth in the 1990s.
It’s worth noting that the rapid price gains, relative to income, paled in comparison with the U.S. housing boom from 2001 to 2006, when prices rose seven percentage points faster than income.
Canada’s price boom was smaller than what happened in the U.S., and initially reflected some catch-up after a long period of weakness that left the market undervalued.
Since 2008, prices have climbed half a percentage point faster than income, though not in a straight line.
After falling 10 per cent in the 2008 recession and softening after several rounds of tougher mortgage rules, prices tended to rebound in response to lower interest rates and steady demand from immigrants and millenials.
Based on the MLS home price index, prices are currently rising about one percentage point faster than personal income, well short of boom-time conditions.
Besides the stimulative effect of rising home prices on wealth and borrowing capacity, the economy was also supported by rapid increases in home building, again after a lengthy period of weakness.
Residential construction rose 28 per cent from 2002 to 2007, adding 0.34 percentage points to annual GDP growth on average.
The economic boost was similar to that in the U.S. from 2001 to 2005.
However, a key difference is that, while U.S. homebuilding was fairly normal in the decade before its boom, Canadian starts were well below average, suggesting the initial boom in construction was a response to pent-up demand.
Based on the above estimates, the rapid increase in home prices and construction added 0.56 percentage points to annual GDP growth from 2002 to 2007, accounting for one-fifth of total growth.
This suggests a moderate correction could have a meaningful slowing effect.
However, to trigger a recession, prices and construction would need to drop sharply.
Based on our model, a 10 per cent decline in prices and construction reduces annual growth by one percentage point, with the two channels contributing about equally.
Given underlying growth of just over two per cent, prices and construction would need to fall more than 20 per cent to spur a contraction.
Since 1980, the most that national house prices have ever fallen is 10 per cent (using quarterly average data), and only during recessions.
Since the housing market does not appear to be egregiously overpriced, a 20 per cent correction is unlikely under normal economic conditions .
Meantime, with homebuilding largely tracking household formations, a downturn in residential construction also doesn’t appear imminent.
Housing starts averaged 175,000 in the first quarter, in line with demographic needs. Aside from the Toronto and Montreal condo markets, there is little evidence of excessive overbuilding across the country.
The bottom line: Our analysis of the impact of the housing boom suggests a moderate 10 per cent correction in prices and construction would slow economic growth by one percentage point, while a steep 20 per cent plunge could pull the economy into recession.
One caveat is that feedback between falling home prices and elevated household debt could make the economy more sensitive to a correction than our simple model suggests. This speaks to the need for households to manage their debt prudently.
On this note, the recent slowing in household credit growth in line with personal income is welcome. Should home prices rise more slowly than income for a period of time, as we anticipate, the economy’s vulnerability to a housing correction would decline further.
Sal Guatieri is a senior economist with BMO.