Canada's Housing Bubble

Aug 19, 2011 Ben Rabidoux theeconomicanalyst.com

When the Bank speaks....people ought to listen!

The Bank of Canada released their Summer 2011 Bank of Canada Review.  Any good economist recognizes the importance of data released by the Bank of Canada as it helps form policy direction.  And that being the case, there are some doozy implications in this most recent report.  In particular, let's zoom in on one report article contained in the report:  "Mortgage Debt and Procyclicality in the Housing Market"

It's not as scary as it sounds.  Here are some key quotes...

On the role of housing in buoying economic activities....and why housing busts really suck!

Boom-bust cycles in real estate markets are a common precursor to banking crises in advanced and emerging economies...the most severe and
costly busts, in terms of lost gross domestic product (GDP), arise when the real estate boom is associated with the increased leverage of households and financial institutions (Crowe et al. 2011). Claessens, Kose and Terrones (2008) show that recessions that coincide with housing busts tend to last longer and be deeper than recessions where no housing bust has occurred. They find that, on average, the cumulative loss to GDP is roughly three times as large when a housing bust coincides with a recession.

This is no shock to readers of this blog.  While I don't see widespread carnage at financial institutions, as their mortgage portfolios are overwhelmingly insured by CMHC, I have often warned that a housing bust would lead to a very nasty recession.  This is because as house prices fall, it buts into consumer spending on two fronts:  1)  It reduces the amount of equity they can withdraw from their home, and 2)  It makes them feel less wealthy, and therefore less willing to spend money on things they don't absolutely need.  This second point is what is known as the 'wealth effect'.  Economists have calculated wealth effect spending at up to 9 cents on every dollar increase in housing value.  This has a major effect on the broader economy as it supports parts of the economy that would not otherwise thrive, leading to lower unemployment, greater confidence, and more demand for housing as people are willing to take on mortgage debt in good times.  But the reverse is also true.  Falling house prices crush spending, leading to higher unemployment, lower confidence, and in turn, lower house prices in a downward cycle.  This is not lost on the authors:

A rising supply of credit, often caused by some form of financial liberalization or technological innovation, is a characteristic feature of a real estate boom. A key mechanism for the boom is feedback between rising house prices and household debt that is created because homeowners are able to use their houses as collateral. Rising house prices increase the value of a household’s collateral and expand the capacity of households to accumulate debt. If new debt leads to further spending on real estate, house prices rise, completing the feedback loop.  The presence of this effect is associated with greater procyclicality in the housing market. In addition, some portion of this increased debt capacity may also finance non-housing consumption. In this case, higher household indebtedness increases the risk of default when income falls during the bust phase of the cycle.

And of course, Canada has seen its share of 'financial liberalization' in the form of falling down payments which were 10% in the late 90s and 20% before that.  Now they are 5%, but even that is not entirely true given that all big banks offer 5% cash back mortgages.  The extension of amortization lengths from 25 to 40....and now back to 30 as of this year, has also aided in the process.  And of course the decision to remove CMHCs maximum insurance ceiling in 2003 had perhaps the greatest impact on debt accumulation as mortgage debt and other consumer debt exploded higher, both in nominal terms and in relation to GDP:

On the growth of Home Equity Lines of Credit (HELOCs):

...Rising house prices allowed existing homeowners to increase their debt levels dramatically (Mian and Sufi 2009b). The authors document that homeowners extracting equity from their homes during the period of rising house prices experienced a jump in default rates as house prices reversed (Mian and Sufi 2009b). This research shows that the feedback effect between house prices and household debt is clearly linked to the degree of vulnerability of the financial system.

I've written repeatedly about the role that rising house prices have played in the boom in home equity withdrawal.  We know what the growth in lines of credit looks like compared to incomes, GDP, and inflation...and it is shocking.

On the role that falling down payment requirements have had on boosting real estate values:

Allen (2010–11) shows that from 1999 to 2004 most households with insured mortgages borrowed up to, or near, the maximum LTV ratio available at the time they purchased a home. Thus, in that period, the typical LTV ratio for a newly issued insured mortgage was in the range of 90 to 95 per cent. This suggests that changes to the maximum LTV ratio could have significant effects on housing markets.

...Together, these studies suggest that the maximum limit on the LTV ratio on a residential mortgage could play a useful role in moderating procyclical movements in house prices and housing market activity.

No kidding.  As I have often said, in 1999, a $10,000 down payment would qualify a borrower for a $90,000 mortgage, provided their incomes supported it.  A year later, that same downpayment qualified the buyer for a $190,000 mortgage.  Do we really need to think hard about how this might affect house prices?

On the role that low down payments have had in buoying consumption:

This part is fascinating.  The author discusses a model that examined the role that down payment had on consumption.  Loan-to-value (LTV) is just the inverse of down payment in this case.  An 80% LTV equals a 20% down payment requirement.  Our current 95% LTV regulations in Canada mean buyers need a 5% down payment.  Here's what they found:

In the case where the maximum LTV ratio is set to 80 per cent, the model responses suggest that a 1 per cent rise in house prices is associated with an increase of 0.1 per cent in consumption, which is close to the estimates of Iacoviello and Neri (2010) for the United States... In the case where the maximum LTV ratio is 95 per cent, the initial response of consumption is three times larger.

Policy implications:

Mark Carney has been VERY clear that he will not use monetary policy as a tool to keep credit demand in check.  It is too blunt a tool.  While they can reign in consumer borrowing by raising interest rates, they will simultaneously crush business spending, a key driver of growth coming out of the recession.  People who have been calling for a rapid rise in interest rates have been dead wrong for exactly this reason.  Last year I was adamant that we would very likely revisit emergency low interest rates before we would see rates shoot to the moon.  Deflation and not inflation remains the greater near-term threat.

With all of that in mind, it is very clear that the Bank of Canada is looking to Ottawa to rein in consumer debt.  They have made some clear policy suggestions in this paper, and you can bet that Jim Flaherty and Stephen Harper will be hearing all about them:

Recent international discussions have begun to examine the merits of adjusting mortgage market rules over time. For example, country authorities could change the maximum LTV ratio in acountercyclical fashion, lowering it during housing booms and raising it when house prices are depressed. One outcome of this type of policy is an increase in the resilience of the financial system since it requires borrowers to have a larger equity stake in their property during booms, thus reducing the potential losses to financial intermediaries during the bust phase when income and house prices fall.  In addition, the lower LTV ratio (higher down payment) would act against the boom in the first place by reducing the extent to which borrowers could extract equity from their homes or take on more leverage to buy a bigger home.

Christensen and Meh (forthcoming) investigate the role of a time-varying maximum LTV ratio in a model based on Christensen et al. (2009).13 They consider the impact when the public authorities respond to a credit boom by lowering the regulatory maximum LTV ratio below its long-run setting of 80 per cent.  The extent of the countercyclical response of the LTV ratio is determined by a regulatory rule that links the change in the LTV ratio to the level of mortgage credit relative to its long-run value.  Housing booms and busts are often attributed, at least in part, to an easing of mortgage-underwriting conditions. We now turn to the case in which lenders themselves supply more credit and consider how the outcome might differ if the LTV ratio was lowered in response.

And with that, the Harper government has just felt the heat to adjust down payments.  This move, coupled with the reinstitution of the regional maximum mortgage ceiling by CMHC would go a long way in letting air out of this exceptionally buoyant market and limiting consequences in the future.

Expect a flood of rebuttals from the mortgage industry...

Cheers,

Ben