Risk Sharing, Flexibility and the Future of Mortgages

Introduction

When people think about the Bank of Canada, the first thing that comes to mind is usually interest rates. But we do have other responsibilities, including issuing safe and secure bank notes—such as the new $10 note, which features Viola Desmond and Winnipeg’s spectacular Canadian Museum for Human Rights, and which was just named the best new bank note in the world for 2018.

Another of our core responsibilities is ensuring that Canadians can conduct their affairs in a safe and efficient financial system. As one of the pillars of that system, Canada’s credit unions share in this responsibility. By safe, we mean a system that is resilient to risks, and where those risks are shared in a desirable way. By efficient, we generally mean competitive, and credit unions bring an important competitive edge to Canada’s financial marketplace.

The forces of competition drive pricing for financial services, but they also drive consumer choice in financial services. More choice means more flexibility and a greater ability to adjust when conditions change. This concept of flexibility is very important to the functioning of our economy, and therefore important to us at the Bank of Canada.

The most important financial decision most Canadians make is the decision to purchase a home and to take out a mortgage in order to do so. After you sign on the dotted line, circumstances can change, posing risks for borrowers, for lenders and for the financial system as a whole. Financial flexibility and optimal risk sharing can help people adapt to changes in their circumstances and help the overall economy adjust to shocks. Recent changes to our mortgage system have brought these issues to the fore in public debate.

In my time with you today, I want to talk about how Canada’s housing markets have been adjusting to economic forces and recent administrative changes. I will then speculate on how our mortgage market could evolve to offer more choice for Canadians, more flexibility for the economy and less risk overall.

Recent developments in housing

We do not have a single housing market in Canada; rather, we have several regional markets. Each has certain national forces in common—such as changes to mortgage lending guidelines and past increases in interest rates—but each also has unique local influences on supply and demand, including provincial and municipal housing policies.

For the past four years we have been following three different housing stories here in Canada. In Alberta and Saskatchewan, we have seen a significant slowdown in housing coupled with falling prices, as the economy works through adjustments to lower oil prices. In sharp contrast, we saw housing booms in Toronto and Vancouver, both of which were experiencing strong population and employment growth and supply constraints on new housing. Low interest rates were of course supporting strong housing demand. Rising house prices fuelled even stronger demand, as first-time homebuyers feared missing out. Foreign buyers were attracted to both markets, adding more fuel. All this helped create extrapolative expectations that prices would keep rising rapidly, a situation that was clearly not sustainable. The third housing story—taking place everywhere else—attracted little attention because it basically looked normal, driven by solid employment growth and historically low interest rates.  

Since early 2016, policy-makers in Ontario and British Columbia have implemented a series of provincial and municipal policies to discourage foreign buying and speculative activity. Policy-makers at the federal level—including the Office of the Superintendent of Financial Institutions (OSFI)—expanded the use of interest rate stress tests on mortgages in both 2016 and 2018. And, with the economy growing strongly, the Bank of Canada began raising interest rates to keep inflation on target. In the wake of all these developments, housing resale activity has fallen sharply in Vancouver and Toronto.

Last month, the Bank published staff research that attempts to untangle the factors behind the huge run-up and subsequent drop in housing activity. It is important to do this because we need to understand well the sensitivity of the economy to higher interest rates. We believe that sensitivity has gone up because of the debt loads people are carrying.

Our research shows that the big rise and fall in housing resale activity in British Columbia and Ontario can mostly be explained by shifts in house price expectations. When house prices are rising rapidly, people tend to extrapolate that experience and buy houses early to avoid further price increases, or to profit from them if they are speculators. In other words, markets become frothy. However, when those price expectations are revised down, demand for houses can cool suddenly. And this is what has happened. The trigger could be anything, including new taxes on foreign buyers, stricter mortgage guidelines, rising interest rates, or simply that rising prices create an affordability roadblock for more and more people.

What we take from this is that it is not higher interest rates and changes to mortgage lending guidelines that have had the greatest effect on housing demand. Rather, it is their interaction with froth that matters most. How much a housing market adjusts depends on how much froth there is. Supporting this conclusion is the fact that many other markets across the country look quite healthy. Resale activity has been solid in places as diverse as Halifax, Moncton, Montréal, Ottawa and, more recently, right here in Winnipeg. This is what we would expect in an economy that is growing, with a rising population and strong labour market, and when interest rates remain very low by historical standards.

Meanwhile, housing markets continue to struggle in Alberta and, to a lesser extent, Saskatchewan. The primary cause appears to be the ongoing process of adjustment to the 2014–15 oil price shock, as well as the renewed weakness we saw late last year. Before 2014, people and businesses were making long-range plans based on continued high oil prices. A lot has changed since then. Most importantly, US oil production has expanded dramatically, causing many to revise downward their long-term expectations for world oil prices. Companies have reduced their investment budgets and worked to reduce their costs. Ultimately, the adjustments are seen in areas that have real consequences for people, such as layoffs, lower wages and lower house prices.

I do not mean to suggest that the new mortgage guidelines have had no effect on housing markets, for they certainly have. The purpose of these changes was to increase the resilience of our financial system in the face of an extremely high debt load. The stress test is designed to ensure that new borrowers can still handle their mortgage payments even if their circumstances change, such as a decline in income or a modest increase in interest rates.

Evidence suggests that the new guidelines have been working as designed. The quality of new loans continues to improve, and fewer mortgages are going to highly indebted buyers. From the house buyer’s perspective, the guidelines look like a cut in purchasing power. But many are adapting by buying less expensive housing or delaying their purchase until they have built up more savings. To the extent that the new guidelines have helped stop the speculative rises in house prices in Vancouver and Toronto, they presumably have also worked to help keep houses from becoming even less affordable.  

We conclude from all this that the fundamentals of the Canadian housing market remain solid, and growth will resume once the effects of reduced expectations for house price inflation and the new mortgage guidelines have been absorbed. Bank staff have also been monitoring the impact of the new mortgage guidelines and higher interest rates on cash flow as households renew their loans. From our initial look at individual loan-level data in 2019, we see that mortgage payments did not rise for most borrowers who recently renewed a five-year, fixed-rate mortgage. Despite increases in our policy interest rate, bond yields and mortgage rates have declined this year along with global interest rates.

The Bank will present its complete analysis of housing and other issues in our next Financial System Review, which will be out in a couple of weeks. We will also post information on our Financial System Hub—a section of our website featuring the latest analysis of key financial issues.

Toward a more flexible mortgage market

Given how different the various regional housing markets can be, it is worth looking to see whether Canada’s mortgage market could become more flexible, giving people more choices and increasing the economy’s ability to adjust.

Canadian homeowners and financial institutions have been well-served by our mortgage market. But it is still important to think about ways to innovate and make a good system better so that borrowers and lenders can make choices that better suit their circumstances.

Of course, there have been innovations in Canada’s mortgage market over the years. Mortgage brokers have appeared on the scene. Financial institutions are doing more business online and reducing the time between application and approval. And mortgages combined with home equity lines of credit, or HELOCs, have become popular. These give Canadians ready access to the equity they have built in their homes, helping people make major purchases and smooth their consumption over time.

Still, there has been significantly less innovation in terms of the actual mortgage product itself. We could look at ways to develop a more flexible mortgage market that gives more choice to customers, lenders and investors, while making the market safer and more efficient.

Diversify mortgage durations

One basic idea would be to encourage more diversity in mortgage durations. It is true that most financial institutions offer fixed-rate mortgages longer than five years. But 45 per cent of all mortgage loans have a fixed interest rate and a five-year term. In comparison, just 2 per cent of all mortgages issued last year were fixed-rate loans with a term longer than five years.

There are historical reasons why the five-year fixed-rate mortgage has been so dominant in Canada. These include legislation from the 1800s that gives borrowers the right to prepay loans after five years without penalty. This provision means that lenders will charge more for longer loans to guard against the risk of a loan being paid back early. And the trend toward five-year loans was reinforced by the Canada Mortgage Bond (CMB) program, which mainly provides five-year funding for financial institutions.

Still, there are compelling reasons why it would be helpful to make more use of longer-duration mortgages. From the consumer point of view, a longer term means they face the risk of having to renew at higher interest rates less often—the longer the term, the fewer renewals take place over the life of the mortgage. Of course, a longer-term mortgage will carry a higher interest rate, but some homebuyers may be willing to pay more to lower their risk. And a longer-term mortgage might not be much more expensive in the long run depending on the details of the loan and the prepayment penalties that apply. At a minimum, we could do more to make people aware of the longer-term options that are available—many people I talk to do not know that longer-term mortgages exist.

As a policy-maker, I see how longer-term mortgages can contribute to a safer financial system and more stable economy. Simple math tells you that of all those five-year mortgages, roughly 20 per cent will be renewed every year. That is a lot of households. If all the mortgages were 10-year loans, only 10 per cent of these homeowners would renew every year.

Another point with longer-term mortgages is that more equity is built up by the homeowner between renewals. This equity position gives the borrower more options at renewal. Therefore, the longer the original mortgage term, the less relevant a mortgage interest rate stress test becomes. In other words, longer-term mortgages shift the risks shared by the lender, the borrower and the system as a whole. All of these dimensions are worth further study.

Develop a private market for mortgage-backed securities

One obstacle to financial institutions offering longer-term mortgages relates to their funding costs. The textbook example of funding is a bank that takes in deposits from customers and turns them into mortgages and other loans. The reality is more complicated. Many institutions seek out wholesale deposits in addition to their regular retail consumer deposits. The biggest banks also get funding by issuing bonds. And, since 1987, Canadian institutions have had the ability to issue government-supported mortgage-backed securities (MBS).

A mortgage-backed security lets institutions package mortgages they have made and sell them to other investors. The investors receive the income from the mortgage payments. The proceeds of the sale become funding that the original institution can use to write more mortgages, or for other purposes. This process is basically invisible to homeowners—they keep making mortgage payments as before. 

Current rules say that only insured mortgages can be used in government-supported MBS. Since 2001, the Canada Housing Trust—which was set up by Canada Mortgage and Housing Corporation (CMHC)—has been buying insured mortgages from institutions to use in the Canada Mortgage Bonds that I mentioned a moment ago. CMBs have been highly successful—more than $230 billion worth of CMBs are outstanding, equal to about 15 per cent of total mortgage debt.

With CMBs and other forms of government support for mortgage financing, the cost for funding insured mortgages is quite low. This has lowered mortgage rates for consumers. However, funding for uninsured mortgages, particularly at smaller banks and mortgage finance companies, is comparatively expensive. It is true that since 2007, Canadian institutions have been able to fund uninsured mortgages by issuing covered bonds. But it can be difficult for smaller institutions to do so because they do not have the same economies of scale as the big banks. What is more, OSFI has capped the percentage of covered bonds that any single institution can have among its assets.

So, it could be helpful for both lenders and borrowers for Canada to develop a private market for mortgage-backed securities. This could be a more flexible source of longer-term funding for uninsured mortgages, particularly those issued by smaller banks, credit unions and mortgage-finance companies. This is an increasingly important point because the market share for uninsured mortgages is increasing, and they cannot be used in CMBs.

Further, CMBs are mostly issued as five-year bonds, matching the dominant five-year mortgage term. If we develop more diversity in mortgage duration, lenders will want different types of funding to help match that. Private MBS could help in this regard.

Finally, private MBS could become another option for investors. Such investors include large institutions like pension funds and insurance companies that need long-term assets to match their liabilities. They also include mutual funds and, perhaps, regular individuals looking for a fixed-income investment that pays more than a regular guaranteed investment certificate.

There is some momentum in Canada toward developing a private MBS market. However, for these securities to be successful investments, they must be designed carefully. After all, many people remember that mortgage-backed securities were at the heart of the sub-prime debacle that preceded the global financial crisis over a decade ago.

A private MBS market in Canada would need to be transparent to give investors more confidence about the quality of mortgages they are investing in. This would allow investors to accept a lower yield, reducing the price difference with regular CMBs. One way to ensure transparency is to develop a public database of mortgages used in securitization. Ideally, this database would include general characteristics about the borrower, property and loan, and data on how the loan has performed over time. An anonymized database would also preserve the privacy of the mortgage holder.

It is worth pointing out that other jurisdictions, including the European Union, the United Kingdom and Australia, already have this type of arrangement in place or are working toward it. We hope that all financial institutions in Canada, from the Big Six banks to the smallest credit union, will co-operate here. After all, such a database could benefit all institutions, not just those that might issue MBS. If institutions can use the information to show how solid their mortgages are, they could lower their funding costs, no matter how they obtain their funding.

New mortgage designs?

The last idea I want to discuss is the design of the mortgage itself. In its most recent budget, the federal government announced the creation of a shared equity mortgage, to be administered by CMHC. The final details will be put in place later this year. This proposal is intended mainly to address housing affordability for first-time homebuyers by giving them an interest-free loan to add to their down payment. The plan will also promote increased housing supply.

Another interesting aspect of the plan is that it should help improve the resilience of the financial system and improve the economy’s ability to adjust to shocks. Although there are many possible variants, this type of mortgage has been analyzed in some depth by two US economists, Atif Mian and Amir Sufi.

Basically, the idea is a rewrite of how the risks around a mortgage are shared between borrower and lender. Because the mortgage is shared, like equity in a company, the risks are shared. Of course, a shared equity mortgage does not make risk go away. The important point is that lenders are better placed than borrowers to carry risk because they can diversify risks across many borrowers. Borrowers are willing to pay to reduce risk, either through a higher interest rate or through sharing their capital gains or losses with their lender. In Canada’s case, with our strong standards for mortgage underwriting, the net result of such a reallocation of risk would likely be a safer financial system.

Of course, there are many other possible variations on mortgage design, so many that it makes me wonder why so little has happened in our mortgage market in my lifetime. I hope that some of the innovative spirit that credit unions have shown in the past will be applied to the mortgage market in the future. After all, one of the founding principles of credit unions is the concept of risk sharing.

Conclusion

It is time for me to conclude.

The Bank of Canada’s mandate is to encourage a safe and efficient financial system—one that evolves to foster resilience, promote flexibility and allow people to make choices that are right for them.

Housing and mortgages are the heart of our system. The Bank of Canada is continuing to watch closely how housing markets are adjusting to the combination of recent provincial and municipal housing policy changes, the revised guidelines for mortgage lending and past increases in interest rates. Some previously frothy markets are still adjusting to a significant shift in price expectations, while other markets appear to be operating in a manner consistent with market fundamentals. As markets stabilize in Toronto and Vancouver, the Canadian housing sector should return to growth overall later this year.

The regional nature of our housing market highlights that all of us should be looking at ways to improve the safety and efficiency of our mortgage market. More choice for borrowers and more ways for lenders to diversify risks are desirable. To be clear, the system is not broken—it has served Canadians and financial institutions well. But we should not stop looking for improvements. And I invite all of you to join this effort.

I would like to thank Ron Morrow, Jason Allen, Brian Peterson, Adi Mordel and Nigel Stephens for their help in preparing this speech.