Forests and Trees: Housing Finance and Macro-prudential Policy in Canada

Speaking Notes for Evan Siddall, President and Chief Executive Officer, Canada Mortgage and Housing Corporation Greater

Bank of England
London, U.K.

Check against delivery

This is a welcome moment for me today personally and professionally. As a friend in New York many years ago, the Governor, who remains a friend today, has long since opened my eyes to a more rewarding career in public service. He helped me see past a mercenary existence amid the trees to a far more fulfilling career attending to the forest. Today is, in a way, an opportunity to thank him for that.

I want to talk to you today about the state of the Canadian housing markets and the use of macro-prudential tools to promote economic growth. I would like to make the case for Canada’s ongoing progress in this area, and suggest that the strength of our financial system and economy are evidence of it.

The Governor and I are both Canadians. We come from a country that emerged from a fragile peace between English and French. Two solitudes originally, we formed a crucible that attracted people from around the world to a place of openness, tolerance and inclusion. Canada is therefore quite naturally an open market, where trading, investment and people from around the world are welcome.

Relatedly, it is notable in a discussion of macro-prudential mechanisms that Canada was the first major country in the world to exit the gold standard and to float its currency. The Canadian dollar has floated for 60 out of the past 66 years. In fact, no other major country has had as much experience with a floating exchange rate.¹

The rebalancing effect of a floating exchange rate has been found to be among the most powerful macro-prudential tools. It keeps policy makers honest and prices differential economic prospects and risks among economies. Few – and possibly no – macro-prudential policies are as effective as unpegged currencies.²

This is Canada’s point of departure in automatic adjustment mechanisms.

Housing Market Vulnerabilities in Canada

Before turning to the specifics of macro-prudential policy in Canada, I should provide a brief overview of the vulnerabilities associated with the housing markets that were instrumental in the most recent macro-prudential changes and that continue to influence our policy thinking.

Concerns about elevated house prices in Vancouver and Toronto are well-known in Canada. CMHC has recently observed spillover effects from Vancouver and Toronto into nearby markets.

These factors were reflected in our Housing Market Assessment, released on October 26. They caused us to issue a “red” warning for the first time concerning the Canadian housing market as a whole.

Worse, the level of household indebtedness in Canada has reached an all-time high and now sits at 168 per cent of income. The Bank of Canada has named this factor the greatest vulnerability to our economy and highlighted growing debt levels among the most vulnerable homeowners as a particular cause for concern. This includes many first-time homebuyers with less job tenure and higher demands on their pocket books.

Less liquid housing investments – compared to other assets – are also hovering near historic highs as a percentage of household net worth, threatening to compound the pro-cyclical effects of house price corrections.

These conditions were begging for a policy response. High levels of indebtedness coupled with elevated house prices are commonly followed by economic contractions. In their book House of Debt, Atif Mian and Amir Sufi called the relationship“ so robust as to be as close to an empirical law as it gets in macroeconomics.”³ The conditions that we now observe in Canada concern us. Increased household borrowing could be jeopardizing our economic future.

Whither Macro-prudential Policy?

It matters to house prices that interest rates are low, of course, providing extraordinary stimulus in a moribund economic environment. Accommodative monetary policy aggravates the vulnerabilities that already exist relative to house price growth and highly-indebted households.

Increasing rates to counteract these effects is too blunt an instrument for addressing specific pockets of imbalance and vulnerability. The International Monetary Fund has concluded that the cost of using monetary policy to “lean against the wind” outweighs the benefits in all but the most exceptional circumstances.⁴ Indeed, macro-prudential tools complement monetary policy by dulling its effect on housing imbalances where it could jeopardize the ultimate goal: economic growth.

And growth must be the objective. My assertion might be obvious, but I think it’s often overlooked. Macro-prudential policy should aim to promote economic growth, not to save banks from themselves, necessarily. Nor is macro-prudential policy a vehicle to manage the business cycle as such. In the end, policy makers must design programs that support the long-term, sustainable growth of the economy. This objective should cause us to focus macro-prudential policy to best target growth.

Macro-prudential Policy in Canada

Returning to Canada, our micro-prudential regime is the responsibility of the Office of the Superintendent of Financial Institutions. OSFI has itself introduced important prudential measures to target elevated household risks. These include, for example, residential mortgage underwriting standards for regulated lenders and mortgage insurers. OSFI has also recently introduced more risk-sensitive capital rules for mortgage insurers, including CMHC. The capital rules for our banks have been further amended regarding exposure to residential mortgages. These are necessary micro-prudential requirements that we have supplemented in Canada via macro-prudential actions.

Canada was an early adopter of a macro-prudential regime, with the introduction of mandatory mortgage insurance 60 years ago, which today is required where a homebuyer has less than a 20 per cent down payment. Mortgage insurance benefits from a federal government guarantee of 100 per cent for CMHC and 90 per cent for our private competitors.

By virtue of its role in guaranteeing mandatory mortgage insurance, the government assumes responsibility to ensure that the programs support a healthy and growing economy. As such, the so-called “sandbox” rules governing mortgage loan insurance have been tightened six times since 2008 in order to manage housing market vulnerabilities. These have applied both to the mandatory homeowner insurance on mortgages above loan-to-value ratios of 80 per cent and also to a voluntary program through which lenders can buy bulk portfolio insurance on low-ratio mortgages. They do this to access CMHC’s securitization programs, which are only available for insured loans.

These sandbox measures have effectively applied macro-prudential limits to insured mortgages, including: reduced allowable amortization periods; specified minimum down payment requirements and debt service maximums; limited mortgage refinancing; a price cap of $C1 million; and required interest rate buffers, or “stress tests,” to be used in underwriting.

Recent Developments

Most recently, on October 3^rd, Canada’s Minister of Finance announced a new round of tightening of these eligibility rules for insured mortgages. In addition to some tax measures, a “stress test” was introduced for all insured mortgages: the Bank of Canada posted rate (about 2 per cent higher than the contract rates at the time of the announcement) must now be used to underwrite all guaranteed mortgages rather than contract rates. This interest rate buffer will specifically help offset the highly stimulative effect of low interest rates. Secondly, while lenders are free to offer more flexible terms for uninsured mortgages, government-backed mortgage insurance will no longer be available for any mortgages, high- or low-ratio, on properties valued above $C1 million, nor those with amortizations beyond 25 years.

We expect that these macro-prudential policy changes will moderate demand for housing in Canada’s housing markets, limiting price increases and making houses more affordable. Finance Minister Morneau also deliberately curtailed the distortionary effects of portfolio insurance, which in my view was stimulating excess credit and contributing to higher levels of household debt. Predictably, this has led to criticism from lenders who had relied on this artificial support for their business models. To quote Warren Buffett, “Never ask a barber if you need a haircut.”

Lender Risk Sharing

At CMHC, we believe that government-guaranteed mortgage insurance is a virtue of the Canadian construct. It instills confidence in our financial system and ensures through-the-cycle stability. And we collect ex ante rents for bearing the burden of a housing crisis tail risk event that political leaders will inevitably need to fund. CMHC has monetized that externality to the tune of over $20 billion in profits and taxes over the last decade.

And yet … do we need such a large presence to preserve the essential virtue of our system? Are we crowding out better ideas through our presence and reactionary defense of the status quo? The proposal to reallocate mortgage risk has been two years in the making⁵ and could represent one of the most significant changes to the Canadian regime since securitization was introduced 30 years ago.

Canada’s Department of Finance recently launched a public consultation on lender risk sharing for government-backed insured mortgages.⁶ This consultation seeks input on a proposal to share a portion of the default risk on insured mortgages directly with lenders, effectively reducing the government’s risk by placing additional private capital ahead of it.

Currently, lenders benefit from a zero risk weighting on guaranteed mortgages. Requiring lenders to have more skin in the game will align interests, reduce moral hazard and allow the system to benefit from enhanced lender risk management.

Two alternatives have been proposed: a deductible or “first loss” approach or a split or “proportionate loss” approach. In the first case, lenders would be responsible for losses up to a fixed amount of the outstanding balance on the loan, with insurers taking on all losses in excess of this limit. This construct is more consistent with the government’s exposure being deeper in the distributional tail. However, it does present different risk profiles between lenders and insurers.

Alternatively, the lender would incur a fixed percentage of the total loss on a loan under the “proportionate loss” model. This construct is more coherent in better aligning risks and better incents lenders to attend to loss severity, or loss-given-default. It would therefore avoid differential approaches to claims adjudication. However, it threatens to add pro-cyclical instability in a crisis, an effect compounded by the fact that severity increases in these circumstances. And since house prices are given to boom-and-bust cycles, this model could undermine our overall objective of supporting economic growth.⁷

The overall impact of the risk-sharing proposal, if implemented, will be to shift risk, capital and rents from mortgage insurers to lenders. At CMHC, we estimate that a modest level of lender risk sharing (such as a 5 per cent first loss or 15 per cent proportionate loss) could increase the typical five-year fixed rate mortgage by a maximum of 10 to 40 basis points, depending on the default risk of the particular mortgage.

We could also see increased pricing differentiation for insurance among borrowers unless we design alternative mechanisms to offset these effects. We are potentially already on this path as a result of OSFI’s introduction of credit scores in calculating mortgage insurer capital requirements. All things being equal, the higher severity of claims we observe in rural areas, in eastern Canada and in our northern regions – which reflect demographics and reduced transaction volumes – could result in higher pricing levels in those areas, particularly under a proportionate loss construct.

Critics have called the proposal "a solution in search of a problem." They cite low arrears rates in a Canada (0.33%) and our experience through the last financial crisis as proof that this proposal represents overzealous policymaking. They don't mention that the Canadian system has not been stressed since the Great Depression. Further, they choose to ignore the strong academic support that loudly warns against the drunken brew of elevated house prices and an advanced credit cycle.

One is reminded of Bertrand Russell's turkey as envisaged by Nassim Nicholas Taleb in The Black Swan: "Its confidence increased as the number of friendly feedings grew, and it felt increasingly safe even though the slaughter was more and more imminent. Consider that the feeling of safety reached its maximum when the risk was at the highest!"

The Future

I am also keen to explore the proper role of government in housing markets through these consultations on lender risk sharing. CMHC currently competes with private insurers. We also offer universal service, insuring Canadians’ mortgages wherever they live – notably in some rural and remote locations that our competitors can opt out of serving. CMHC has deliberately reduced the scope of our own commercial activities: our estimated market share in the mortgage insurance industry has declined from a high of close to 90 per cent in 2009 to about 56 per cent today.

Some interesting alternatives to the status quo bear further exploration through these consultations as well. Rather than offering a whole life policy, guaranteeing 100 per cent of the mortgage for the length of its life, should insurance end at a loan-to-value floor? Separately, could Canada serve as a laboratory to explore the merits of novel ideas like Mian and Sufi’s proposed shared-responsibility mortgages?

And finally, while we’re talking about risk sharing, a conversation about collateral is in order. We know that one of the most influential factors driving insurance losses is the loan-to-value ratio. In fact, it exerts a double whammy: it affects both the probability of default and the severity of the loss. Further, research by Yale economist John Geanakoplos suggests that low down payments play a central role in escalating house prices.⁸ Politicians are tempted to help first-time homebuyers enter the market, but low down payments may be part of the problem adding to affordability pressures and macro-economic vulnerabilities.

In 1992, the Canadian government reduced the minimum down payment from 10 per cent to 5 per cent for first-time homebuyers, in order to provide economic stimulus. This benefit was extended to all homebuyers in 1998. Coupled with the personal exemption from capital gains taxes on the sale of principal residences and other programs, Canadians have very powerful incentives to own homes. At 69 per cent, our homeownership rate is among the highest in the world. While homeownership has been an effective vehicle of forced savings and retirement security, it may also constrain labour mobility.

At CMHC we say our mission is to “help Canadians meet their housing needs.” I have yet to be convinced that people in our country “need” access to 19:1 leverage to buy homes. In fact, it may be a fool’s bargain with the extra demand simply feeding higher house prices: the benefits of the policy accruing to wealthier home sellers rather than to the young first-time homebuyers it purports to help.

While additional measures are not currently on the table, we will continue to be vigilant and monitor the housing market – and share our views as a policy advisor to Government – in order to ensure the long-term stability of the market and of the financial system.

Relatedly, I think the objective of supporting housing affordability demands that CMHC explore a potential future path to higher minimum down payments. Our federal government took an initial step in this direction last December, requiring 10 per cent contributions on the portion of house prices above $C500,000. At a $C1,000,000 house purchase price – the maximum we will insure – the minimum down payment for an insured mortgage is now 7.5 per cent. Some provincial governments are considering support for first-time homebuyers; however, I fear these measures may only further increase demand and therefore house prices.

In a low-for-long environment, it is also worth exploring the future merits of a loan-to-income limit, which the U.K., Ireland and others have introduced. An LTI cap would augment our existing debt service limits on the portion of income spent on housing, and further limit the effect of low interest rates on housing demand.⁹

As contributors to policymaking, it is CMHC’s job to keep an eye on the forest and prepare options for the future. Measuring our success will be complicated by the lack of a counterfactual. Thus far, however, our GDP growth path provides tentative evidence that we could be on track.

Closing

I’d like to close on a word of caution and humility, and remind ourselves of the risks of being evangelical about macro-prudential policies.

My late father was a fan of Ogden Nash, the whimsical American poet. I came across one of his books as I was preparing my remarks. Lamenting the fact that billboards had blocked the view of forests, I’d say that Nash also – by analogy – issues a Minsky-esque warning about being over-confident in erecting protective macro-economic barriers:

I think that I shall never see
A billboard lovely as a tree.
Indeed, unless the billboards fall
I’ll never see a tree at all.

Effective micro-prudential regulation remains our first line of defense. Let’s not lose the trees for the forests. 

Thank you.

¹ Gordon Thiessen, “Why a Floating Exchange Rate Regime Makes Sense for Canada” (speech at Montreal Board of Trade, Montreal, 4 December 2000).

² Atif Mian, Amir Sufi and Emil Verner, “Household Debt and Business Cycles Worldwide,” Kreisman Working Paper Series in Housing Law and Policy. Paper 41, January 2016.

³ Atif Mian and Amir Sufi, House of Debt (University of Chicago Press, 2015).

⁴ International Monetary Fund, “Financial Stability and Interest-Rate Policy: A Quantitative Assessment of Costs and Benefits,” IMF Working Paper, WP/16/73 2016.

⁵ Bank for International Settlements, Basel Committee on Banking Supervision, “Mortgage insurance: market structure, underwriting cycle and policy implications,” August 2013.

⁶ Finance Canada, “Balancing the Distribution of Risk in Canada’s Housing Finance System: A Consultation Document on Lender Risk Sharing for Government-Backed Insured Mortgages,” 21 October 2016.

⁷ Richard Herring and Susan Wachter, “Bubbles in Real Estate Markets,” Zell/Lurie Real Estate Center, Working Paper #402, March 2002.

⁸ John Geanakoplos, “Liquidity, Default, and Crashes: Endogenous Contracts in General Equilibrium,” in M. Dewabtripont, L. P. Hansen, and S. J. Turnovsky, eds., Advances in Economics and Econometrics: Theory and Applications (Eighth World Conference, Volume II, Econometric Society Monographs, 2003, pp. 170-205).

⁹ Recent Bank of Canada modelling indicates that highly indebted households are even more at risk with interest rates near the zero lower bound. See Carolyn Wilkins “(S) low for Long and Financial Stability” (Official Monetary and Financial Forum City Lecture, 14 September 2016).