Standing Committee on Finance
House of Commons
Ottawa, ON
K1A 0A6

Attention: Mr. Wayne Easter, P.C., M.P., Committee Chair

Dear Committee Members,

I am writing in lieu of an in-person appearance at your Committee, which we were unable to arrange. My intention is to provide the public policy rationale behind the mortgage interest stress test, which is the subject of some misunderstanding. I hope my comments assist the Committee in its important work on behalf of Canadians. We will also publish this letter on our website.

Two Stress Tests: Macro- and Micro-prudential

There is some confusion over the merits and intentions of the mortgage interest rate stress test, which was first introduced in 2010 by then Minister of Finance Jim Flaherty, who explained that it was intended “to help prevent Canadian households from getting overextended.” Incidentally, the idea for this policy was suggested by my predecessor, Ms. Karen Kinsley. It applied initially to insured mortgages (down payments below 20%) having variable interest rates or terms of less than five years.

However, the fact that only 30% (using 2016 data) of new insured mortgages were variable rate or shorter than five years limited its effectiveness. In October 2016, Minister Morneau extended the applicability of the stress test to five-year mortgages. In protecting the economy and younger households from the burden of excessive borrowing, the result was to defer some people’s ability to buy a home. We expected a one-time reduction of up to 25-30% of insured mortgage volumes, partially offset by families’ ability to find other sources of borrowing or to buy smaller homes. While it is hard to disentangle other effects on housing activity, CMHC’s volumes declined by 11% for the first three quarters of 2017 to the same period in 2018.

In quite a separate measure, OSFI subsequently amended its B-20 Guideline to expand and strengthen the stress test for uninsured mortgages (down payments of more than 20%), effective January 2018. OSFI was exerting additional influence on lenders’ underwriting prudence for less risky low ratio mortgages. Since these less leveraged households will generally have more money, they also have alternatives — and the impact of the B-20 stress test per se has been less acute.

There is an important distinction to be made between the two stress test regimes. The Minister of Finance’s stress test on insured mortgages is macro-prudential economic policy and OSFI’s B-20 stress test is micro-prudential. While the value of the stress test is apparent from the fact that it kills two birds with the same stone, the intentions are quite different. OSFI’s rationale is founded on prudent underwriting standards, which do not change with interest rates, house prices nor the economic outlook. Macro-prudential rules are targeted at preserving financial stability.

These policies have been wrongly accused of having “unintended consequences.” Public servants spend a great deal of time evaluating consequences and how behaviours might change as a result of policies in normal times and in times of financial instability — a state that is too often ignored. The call by some people for relief from the so-called “unintended consequences” of the stress test for mortgage renewals provides a good example. At first glance, this seems unfair.

Critics are however overlooking the fact that underwriting standards must apply in both good times and in bad. In a February 2019 speech, OSFI Assistant Superintendent Carolyn Rogers explained why renewals should not be exempted from the stress test. In brief, OSFI doesn’t want to stimulate competition among banks for weaker credits.

We can take the thinking a step further. Lenders often engage in pro-cyclical behaviour such as dumping assets in the face of a crisis, especially weaker assets, sending asset values even lower. We wouldn’t want lenders hiding behind the stress test to justify refusing to renew mortgages. OSFI evidently chose a lesser evil: effectively requiring that existing lenders support their existing customers — a practice they will monitor — in a crisis. Inherently, the policy means that lenders must therefore re-price these exposures instead of refusing to renew.

Easing of the stress test buffer is therefore not called for as rates rise. In neither case is an interest rate shock the underlying concern. OSFI will presumably want to maintain an underwriting buffer through the economic cycle. And we will continue to support it for insured mortgages until debt to income ratios moderate significantly in Canada. The single largest risk to an insured mortgage is the homeowner’s  unemployment — not higher interest rates. Our objective is to protect economic growth and jobs.

The Financialization of Housing and Household Debt

House prices tend to track growth of disposable income in the long term. Over the past 20 years, average Canadian house prices have increased by 3.8% per year while national income (GDP) has increased in real terms by 1.9% per annum. Gross household debt has increased from $539 billion in 1998 (106% of disposable income) to $2.2 trillion, or 178% of disposable income in the fourth quarter of 2018. Said otherwise, household debt has increased over 20 years from 58% to 99% of GDP and mortgage debt from 37% to 65%.

Our appetite for housing, both as shelter and as an investment, has been filled by borrowing — a liquidity boom akin to the historical cases I mentioned earlier.

Debt is a claim on future income. Borrowers monetize future income to buy houses today. Since that future income must therefore be dedicated to repaying debt, it can’t be used for consumption — buying goods. And that means that the consumption on which our future economic prosperity rests (58% of our GDP) is already spent. Our future spending is sustained only if we borrow more, such as by using our homes as bank accounts through home equity lines of credit.

As policy makers, we have a choice. We can continue to fuel this liquidity or we can gently let the pressure leak out. The stress test was designed to do exactly that.

The Stress Test and Protecting Economic Growth

The mortgage stress test is exactly the kind of policy we need to protect our economy. Minister Morneau was clear in his October 2016 announcement regarding the stress test for insured mortgages: the parameter was being added to protect our long-term economic health. This rationale was reiterated in Budget 2019:

[The stress test has] contributed to slower growth in house prices and reduced speculation in key areas, helping to limit the amount of debt Canadians must take on to buy a home and improve housing affordability. Nonetheless, household debt remains high, and there continue to be risks in the global economic environment.

In their book on the US Financial Crisis, called House of Debt, two leading economists showed that the combination of high house prices and elevated debt led inexorably to reduced future economic growth. They called it “as close to a rule as exists in macroeconomics.”

The Bank for International Settlements published a study in 2017 that showed that while borrowing resulted in a temporary bump in consumption — lasting less than a year, borrowing above 60% of gross income were associated with longer-term negative impacts on consumption. Moreover, borrowing above 80% of gross income intensified the drag on GDP growth. Canada is well above both thresholds.

The potential consequences of our debt-fuelled real estate boom in Canada are serious. Asset bubbles fuelled by too much liquidity create the conditions for their own demise. When this money is borrowed and the investment value turns south, the vortex that the sudden withdrawal of that money creates a panic. Asset values crash until cooler heads prevail. Whether tulip bulbs in 17th Century Netherlands, the South Sea Bubble in 18th Century Britain, the 1920s stock market bubble, or early 21st Century US real estate, the pattern is foretold. Just ten years after such a crash, we have fallen into the “this time is different” trap of complacency.

We have a responsibility to prevent these tragedies. And while I’m not predicting, it nonetheless could happen here: of the 46 banking crises for which we have housing data, over two-thirds were preceded by real estate boom and bust cycles. Moreover, we must avoid policies that serve only to enrich the people who have already made tax-free gains on real estate; rather, we need to be deliberate to avoid further inflating house prices. Housing affordability is CMHC’s raison d’être and we will consistently argue for that.

The Real Culprit: High House Prices

Critics of the stress test ignore the fact that high house prices are the overwhelming reason why home ownership is out of reach. There are many economic and other factors behind this. In fact, the stress test has helped moderate house prices, making home ownership easier. If I can borrow an analogy from my colleague, OSFI Superintendent Jeremy Rudin: having braked hard enough to stop your car from barreling down a hill, you don’t declare victory and release the brakes.

We must of course attend to the people at the margin who bear too much of the cost. This is exactly the aim of the First Time Home Buyers Incentive. And that program has been scaled to have a near insignificant impact on prices as well as to avoid stimulating borrowing.

It’s worth comparing the FTHBI to the two alternatives promoted by some people in the real estate industry. In particular, the Canadian Home Builders Association, the Mortgage Professionals Association of Canada and the Ontario Real Estate Association have all called on the government to ease the stress test and restore 30-year mortgage amortizations. Either of these would have added to housing demand across the board (not just targeted to a small segment) and price inflation of as much as 1-2% in our large cities.

Both ideas would also stimulate increased borrowing. Paul Taylor of the MPC was bold enough to criticize the FTHBI to this Committee last week in suggesting that we should actually encourage people to borrow 4.8 times income. This same rhetoric led successive US administrations to promote home ownership to a fault. Apparently, the MPC is content to see home builders, real estate agents and mortgage brokers receive short term benefits while Canadians bear the long-term costs.

My job is to advise you against this reckless myopia and protect our economy from potentially tragic consequences. Indeed, it was my primary aim in leaving a private sector career in finance for public service: to help prevent a repeat of the harm that excessive mortgage lending created for hundreds of thousands of households just a decade ago.

The stress test is doing what it is supposed to do. Please look past the plain self-interest of the CHBA, MPC and OREA and see house price moderation as helpful: an intended consequence. Choose instead to heed the consistent views of those of us who are unconflicted: the Department of Finance, OSFI, the Bank of Canada, the IMF — again just this week, and CMHC. We will of course continue to monitor housing markets to ensure that the stress test and all sandbox measures are having the desired effects and the Minister of Finance or OSFI may make changes in the future.

CMHC’s analysis concluded that since first introduced in 2010, changes in stress test requirements since 2010 have helped reduce house prices nationally by 3.4% versus where they otherwise would have been. Correspondingly, in an astonishing piece of work, economists at TD Bank argued that the stress test should be removed so that house prices can increase by $32,000. CIBC’s economist has also called for a re-assessment of the stress test. Since a federal government guarantee stands behind lenders’ insured mortgages, these appear to be cases of evident moral hazard. I doubt they’d be as cavalier if it were their risk.

In closing, therefore, I would ask you to see past those who insist that everything will unfold benignly, as we all nonetheless hope. As a former US Senator once said, ”A billion here, a billion there; sooner or later it adds up to real money.” And then we have a problem — and one we could have avoided.

I regret that I was unable to deliver these messages in person and respond to questions. Thank you nonetheless for the opportunity to write to you and for your diligent work on behalf of Canadians.

Yours truly,

Evan Siddall
President & CEO
Canada Mortgage and Housing Corporation

Explore related content using the tags below:

Date Published: May 23, 2019