Stephen S Poloz: Year-end economic progress report - financial vulnerabilities in focus

Remarks by Mr Stephen S Poloz, Governor of the Bank of Canada, to the CFA Toronto, Toronto, Ontario, 6 December 2018.

The views expressed in this speech are those of the speaker and not the view of the BIS.

Central bank speech  | 
06 December 2018


Today we are combining two traditions-one old, one new. The older tradition, which goes back more than a decade, sees the Bank of Canada Governor talk about the state of our financial system in a year-end speech in Toronto. The newer tradition, which started earlier this year, is to offer an Economic Progress Report four times a year, following the interest rate announcements that do not come with our Monetary Policy Report (MPR), like yesterday's.

My motivation for combining these two traditions is more than just trying to raise productivity. Ten years ago, the global financial crisis gave us the preconditions for a repeat of the Great Depression of the 1930s. Thanks to aggressive fiscal and monetary policies, we avoided that. But a decade of massive monetary stimulus has brought us to a critical phase in the economic cycle.

The Canadian economy has been operating near its capacity for over a year now, unemployment is at its lowest in decades, and inflation is on target. Historically, this is the point in the business cycle where inflation pressures can start to build, so it is natural that we are looking to move interest rates to a neutral level. At the same time, the decade of exceptionally low interest rates has led to a buildup of household debt, consisting mostly of mortgages. And we are all aware of large increases in house prices in recent years, particularly in some of our largest cities.

These financial vulnerabilities have made monetary policy more complicated. Understanding them and integrating them into our policy process has been a priority for the Bank. Meanwhile, the world has not stood still since our last MPR in October. Indeed, there have been more macroeconomic developments than usual. My focus today, then, is to bring you up to date and to set out how financial stability concerns are captured in our decision-making process.

The growth of financial vulnerabilities

Let me start with a few words about how our financial vulnerabilities became so large to begin with. Fact is, debt builds up during the recovery phase of every economic cycle. Low interest rates stimulate a recovery by encouraging households and firms to borrow. Normally, the economic cycle is short enough that the risks posed by these imbalances are relatively minor.

Of course, the past 10 years have been far from normal. Interest rates have been extraordinarily low for an extraordinarily long time. The inevitable result has been strong demand for housing, rising house prices and an accumulation of household debt of historic proportions. The associated buildup in financial vulnerabilities has been a preoccupation of the Bank for several years. Indeed, household indebtedness in Canada was rising strongly long before the global financial crisis, mainly as a result of an easing of mortgage rules some 15 years ago, financial innovation and declining interest rates.

In response to these developments, the federal government has implemented a series of more restrictive macroprudential policies-designed to ensure that new borrowing is safer. In particular, with the revised B-20 guideline that came into effect earlier this year, all new mortgage borrowing from federally regulated institutions became subject to some form of stress test to ensure borrowers can handle an increase in interest rates. You can find a lot more detail about this and other stability issues on our Financial System Hub on the Bank's website.

Now, I hear all the time from people who say this stress test should not apply in housing markets that have not been overheating. But the point of the change to B-20 was not to cool housing markets. The point was to increase the future resilience of new household debt. All Canadians face the risk of higher interest rates, not just those in hot housing markets.

Still, there is no doubt that the combination of these stress tests, higher interest rates and housing policies put in place by provincial and municipal governments are affecting household borrowing. We are seeing far fewer mortgages being taken out at debt-to-income ratios above 450 per cent. And credit data show mortgage borrowing growth has slowed this year to a rate of just over 3 per cent. At this rate, the aggregate debt-to-income ratio is likely to establish a moderate downtrend.

In setting monetary policy it is crucial to disentangle the impacts of these various policies. We need to know how much of the slowdown in credit growth is due to higher interest rates, rather than to the other policies. How large is the impact of higher interest rates on household spending? And how are various macroprudential policies, including the B-20 guideline, affecting the behaviour of borrowers and home builders?

Clearly, there is no way to get precise answers to these questions. But we need to understand how developments in the financial system affect the real economy and the risks it faces. I spoke on this topic almost four years ago during a lecture at Western University. I talked about the idea of developing a grand synthesis-a utopian economic model capturing how the financial system would affect the real economy and vice versa, and helping us deliver on our inflation-control mandate.

We may never reach that utopia. But the good news is that we have made significant progress. Let me mention three important improvements.

First, a new concept that was developed at the International Monetary Fund-growth at risk-is now being used by Bank staff to help us understand the links between the financial sector and the real economy. It captures the rise in financial vulnerabilities and the downside risks to economic growth associated with them. Accordingly, when contemplating a change to interest rates, we can estimate both the usual direct effects on the economy, and the indirect effects through the financial vulnerability channel. The growth-at-risk framework is not the grand synthesis, but it does give us a more rigorous way of thinking about financial vulnerabilities within our risk-management policy framework.

Second, we have upgraded our main economic model to incorporate household debt accumulation and therefore to capture the fact that the economy is more sensitive to interest rate movements when debt levels are high. It also embraces the link between debt accumulation and rising house prices. These are important steps to help us understand the impact of these vulnerabilities on the economy.

Third, Bank staff are working with new sources of microdata to deepen our understanding of how higher interest rates affect mortgage holders. We now have access to anonymized data at the individual loan level dating back to 2014, covering roughly 85 per cent of mortgage borrowing over this period. This includes information on mortgage size, household income, the interest rate at origination, the mortgage term and the amortization period.

These data allow us to calculate how households are being affected by higher interest rates through the mortgage renewal cycle. By estimating the effects of higher interest rates on monthly household expenses, we can forecast how spending on other purchases will be affected, too.

So far, this enhanced framework has performed well for our forecasters. However, it is important to remember that, to date, most households have been renewing at mortgage rates that are pretty similar to the rates they signed up for five years ago. As we go forward, people will increasingly face higher interest rates when they renew, and we will learn more about how people are adjusting.

It is clear that many highly indebted households will face a difficult adjustment as their mortgages reset and interest rates go up. Still, these adjustments will be far less demanding than if there were a serious negative economic shock, especially if financial vulnerabilities were allowed to continue to grow unchecked.

Now, the fact that household spending is behaving roughly as expected gives us more confidence that we understand what is happening in the economy. But borrowers and lenders are continuing to adjust to rising interest rates and the new mortgage rules. So, we are closely watching a number of trends in mortgage markets. For example, the share of mortgages that originate outside federal jurisdiction, including from private lenders and credit unions, is rising. These borrowers are not subject to a formal interest rate stress test. We have seen a rise in the share of mortgages issued by private lenders in the Toronto area, although we do not have this type of data for cities outside of Ontario.

We have also seen a greater share of highly indebted borrowers taking out variable-rate mortgages. In doing so, they are lowering their debt-service burdens because, usually, the interest rate on a variable-rate mortgage is lower than on a fixed-rate mortgage. This frees up money for spending or saving in the short term, but exposes the borrowers to unexpected increases in interest rates down the road. That said, the stress test in place gives us confidence that these borrowers can manage significantly higher payments, if need be.

In summary, while the quality of new lending has improved, the stock of risky mortgages remains high. Over time, those mortgages should become less risky as they are slowly paid down. Still, this vulnerability will persist for many years.

Closely related to the buildup of household debt are developments in housing markets. Everybody is talking about this issue-which is no surprise when you see that house prices in the Toronto area are close to 40 per cent higher than they were three years ago. In the Vancouver area, the increase has been even larger-about 50 per cent. Outside of those two areas, the average home price has increased just 5 per cent over the same three years.

To be clear, fundamental factors have pushed up house prices in Toronto and Vancouver. Strong population and employment growth has supported housing demand. The cost of several inputs, such as construction labour and development fees, has also increased. At the same time, various policies and other factors have limited supply growth in both places. If supply does not expand in a climate of strong demand, you have a recipe for rising prices.

That said, it seemed clear to us that the price growth was being magnified by speculative activity, particularly during 2016-17. Some buyers were accelerating their purchases, motivated by fear that they would be priced out of the market if they waited. Others, primarily investors, were buying real estate on the assumption that prices would keep rising. This is significant, because when speculative activity drives prices unsustainably higher, an economic shock can prompt a sharp decline. Anyone who remembers the housing market of the early 1990s in Toronto and Vancouver will recognize this point. And the impact of such a drop is magnified when the homeowners are highly indebted.

The Bank has raised the key policy interest rate five times over the past year and a half, by a total of 125 basis points. And I have heard from some Canadians, more so lately, who are concerned about the impact of these rate increases on housing affordability. However, given the combined impact of provincial and municipal housing measures and tightened macroprudential policies-not to mention higher interest rates-house prices for Canada as a whole are now growing at an annual rate of roughly 2 per cent. It seems to me that this slowdown in house price inflation is much more meaningful in terms of affordability for first-time homebuyers than the interest rate movements we have seen. The basic laws of economics tell you that measures to increase supply would be the most effective way to support affordability. And measures that increase demand, with no corresponding increase in supply, could make housing affordability worse.

Speaking of the uptrend in interest rates, a related risk to Canada's financial stability that we have been watching closely is that of a snapback in global interest rates. Because Canada would typically import some 60 to 70 per cent of any rise in global bond yields, we would see the attendant effects on our mortgage rates, even if the Bank of Canada policy rate was held unchanged. This risk remains top of mind, particularly given bond and equity market volatility in recent weeks. Most observers would contend that the likely catalyst for such a risk would be some inflation surprise coming from the United States, the probability of which has risen with the response of the US economy to fiscal stimulus. However, our outlook remains that the US economy will moderate to a more sustainable pace next year and into 2020 and that inflation expectations remain well anchored.

So, let us sum up the financial stability situation. Governing Council judges that the overall level of risk to the Canadian financial system is about the same as it was six months ago, when we published our Financial System Review. New mortgage borrowing is more sound, and house price growth has decelerated. Nevertheless, the stock of household debt will stay high for years, and house prices remain elevated in certain markets. Our new growth-at-risk framework shows clearly that macroprudential policies have worked to mitigate financial stability risk, thereby improving the risk-management problem faced by monetary policy. 

Macroeconomic risks and the inflation outlook

Let me now turn to the macroeconomy. As we said in October, the economy has been operating close to its capacity for more than a year and inflation is on target. Since October, there have been a number of important developments.

First, there have been growing concerns of a global economic slowdown. I would note that our forecasts were already calling for a moderation of economic growth in 2019-20, but this would only bring us to a sustainable growth track and would not be cause for concern. Nevertheless, the main risk we see to that outlook today is ongoing trade tensions between the United States and other countries, particularly China.

Rising tariffs will slow economic growth and reduce productivity on both sides, and will raise inflation risks besides-a combination we used to call stagflation. This combination is particularly challenging for monetary policy, since it forces a trade-off between cushioning economic growth through lower interest rates and containing inflation risk with higher interest rates. Because the effects on the economy would likely prove to be structural, rather than cyclical, I have to believe that containing inflation risks would become paramount in an outright trade war.

Importantly, the risks around global trade are two-sided. Yes, there is growing evidence that trade actions are already having negative macroeconomic effects. But as central bankers we cannot focus only on the worst-case scenario. The upside risk is that the United States and China come to terms, and the global economy enjoys a new source of lift. Events this past weekend in Buenos Aires were somewhat encouraging on this front. So, we continue to weigh both sides of the issue.

In terms of the Canadian economy, it is fair to say that the data released since our October MPR have been on the disappointing side. While the GDP data for the third quarter were close to our expectations overall, the underlying composition of growth was not, and the economy has less momentum going into the fourth quarter than we believed it would.

While recent data from the housing sector have been softer than expected, we believe this is the result of a significant adjustment in new construction from single dwellings to multiples, and this adjustment has prolonged the slowdown in housing construction that began earlier this year. Population and employment growth, and therefore fundamental demand for housing, remain strong. Credit growth has also steadied, and all of this supports our view that the market is stabilizing.

Business investment unexpectedly declined in the third quarter. We had identified earlier this year the likelihood that uncertainty over the future of NAFTA would hold back investment decisions. It now appears that this effect was very strong during the summer, when uncertainty was at its peak. The other big factor restraining business investment was the delay in the Trans Mountain Pipeline project.

The signing of CUSMA is likely to support a rebound in investment, particularly given ongoing capacity constraints, although governments still need to implement the agreement. Furthermore, the federal government's recently announced tax changes will lead to a further strengthening in investment. This would also suggest continued growth in exports, which have been supported by strong foreign demand, but constrained by tight capacity.

In its latest GDP release, Statistics Canada revised downward its historical estimates of economic growth. Most of these revisions pertain to shifts in economic structure beginning in 2015, as the economy was adjusting to steep declines in oil prices. It may seem odd that developments in 2015 could still be affecting our view of the economy in 2018, but they do. The level of GDP today is now believed to be nearly 1 per cent lower than previously thought. The effect of this revision on the inflation outlook will depend on how much of that shift is in demand and how much is in supply or economic capacity, and therefore how our estimates of the gap between the two are affected. We will say more about this in our January MPR, once our analysis is complete.

Much of Governing Council's discussion was focused on oil. Global oil prices are well below our forecast assumptions made in our October MPR, primarily due to supply forces. There is also an important overlay of worry about moderating global economic growth, given heightened trade tensions, with implications for future oil demand.

The main source of additional oil supply is the same as it was back in 2014: the United States. For reference, the world consumes about 100 million barrels of oil per day. Back in 2008, US production was about 7 million barrels per day. By 2014, the shale revolution had pushed that figure up to 12 million. Today, US production is over 15 million barrels per day, more than double the 2008 level. That is 2 million more than just a year ago, and US exports of oil have risen by the same amount over that time.

These US developments dwarf Canadian production-we produce about 5 million barrels per day and export just over 3 million. Weak global oil prices have a direct impact on Canada, as we well know from our experience in 2015-17. This affects all producers, in the east and west. Although the Canadian economy overall had largely completed its adjustment to lower oil prices by mid-2017, adjustments to cost structures, and to wages and employment in oil-producing regions, have continued. Indeed, the share of oil and gas production in the Canadian economy has declined since 2014, from about 6 per cent to about 3½ per cent today.

The recent decline in world oil prices has been magnified in Western Canada, as the discount applied to our heavy oil, Western Canada Select, widened to a record. This is largely due to transportation constraints, but was compounded this autumn by maintenance shutdowns in some key US refineries. Since our October rate decision, the heavy oil discount has declined somewhat as US refineries have come back on line. However, in recent weeks, the discount has also spread to the price for Western Canadian light crude oil, Edmonton Par, as stored inventory has hit record levels. Alberta's initiative to impose output reductions and add more rail capacity will help clear the backlog, and more pipeline capacity would certainly help in the long term.

While we did not prepare a completely new economic forecast for yesterday's decision, we will do so between now and our next decision in January. It is already clear that a painful adjustment is developing for Western Canada, and there will be a meaningful impact on the Canadian macroeconomy. That said, given the consolidation that has taken place in the energy sector since 2014, the net effects of lower oil prices on the Canadian economy as a whole, dollar for dollar, should be smaller than they were in 2015.

Summing up, then, as I said, a lot has happened since our October MPR. But let us not forget that these developments have come against a backdrop of an unemployment rate at a 40-year low and inflation close to target, consistent with an economy that has been operating close to its capacity. We will assess all of these new developments in our new projections in the January MPR.

Governing Council determined yesterday that the current level of interest rates remains appropriate for the time being. And, weighing all of these developments, we continue to judge that the policy interest rate will need to rise into a neutral range-somewhere in the neighbourhood of 2.5 to 3.5 per cent-in order to achieve the inflation target. The pace at which this process occurs, of course, will remain decidedly data dependent. We will continue to gauge the impact of higher interest rates on consumption and housing, and monitor global trade policy developments. The persistence of the oil price shock, the evolution of business investment and our assessment of the economy's capacity will also factor importantly into our decisions about the future stance of monetary policy.


It is time for me to conclude. I hope my speech has shown you how financial vulnerabilities fit into our monetary policy discussions. We have made progress in thinking about these and other vulnerabilities and understanding the risks they pose to the economy. I've also tried to give you a sense of the multiple issues we are wrestling with in managing the macroeconomic risks facing the Canadian economy. We will continue to manage these risks as we pursue our mandate to control inflation and promote the economic and financial welfare of Canadians.

I wish you all a very happy holiday season and all the best in the new year.

I would like to thank Don Coletti for his help in preparing this speech.