Stephen S Poloz: Canada's economy and household debt - how big is the problem?

Remarks by Mr Stephen S Poloz, Governor of the Bank of Canada, at the Yellowknife Chamber of Commerce, Yellowknife, Northwest Territories, 1 May 2018.

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

Central bank speech  | 
03 May 2018
PDF version
 |  5 pages

I would like to thank Jing Yang for her help with this speech.


Shakespeare wrote, "Neither a borrower nor a lender be." Well, that may have been reasonable advice back in Hamlet's day, but it is hard to imagine a modern economy like ours functioning under that dictum.

For most Canadians debt is a fact of life, at least at some point. Borrowing can help someone get a higher education, or buy a new car, or purchase a home. Simply put, debt is a tool that allows people to smooth out their spending throughout their life.

The amount of debt held by Canadian households has been rising for about 30 years, not just in absolute terms but also relative to the size of the economy. At the end of last year, Canadian households owed just over $2 trillion. Mortgages make up almost three-quarters of this debt.

While debt is indispensable for our modern way of life, it has been a growing preoccupation for the Bank of Canada for several years now. That is because high debt levels can make us vulnerable to negative events-individuals as well the entire economy.

There are two ways to look at this. Traditionally, our focus has been on the vulnerability of Canada's financial system arising from elevated indebtedness. This means analyzing how our banks would manage a serious economic recession with high unemployment and increasing debt defaults. But the Bank is also focused on the vulnerability of our economy to rising interest rates, given high household debt. There is little doubt that the economy is more sensitive to higher interest rates today than it was in the past, and that global and domestic interest rates are on the rise.

So, today I want to talk about household debt in Canada-the dynamics that led to its buildup, how big a problem it is for Canadians now, and how we can manage the risks in the years ahead.

How did we get here?

Two trillion dollars of debt is a big number. Let us try to put some context around it. A common way to measure household debt is to compare it with the amount of disposable income people have. In Canada's case, household debt is around 170 per cent of disposable income. In other words, the average Canadian owes about $1.70 for every dollar of income he or she earns per year, after taxes.

That ratio is a Canadian record, and up from about 100 per cent 20 years ago. Although this ratio is on the high side, other economies such as Sweden, Norway and Australia have even more household debt relative to disposable income.

This international comparison reveals some common factors. Like Canada, the countries I just mentioned have all seen decades of steadily rising house prices. They all have high rates of homeownership and deep, well-developed mortgage markets. Like Canada, mortgages in Australia are typically amortized over 25 to 30 years. In Norway and Sweden, you can find mortgages where the homeowner is only making interest payments, and the principal is passed on from one generation to the next.

Aspiring to own a home is part of our culture. It is also a way to build wealth for the future, as house prices have tended to rise faster than incomes. My colleague, Deputy Governor Larry Schembri, took an in-depth look at the drivers of house prices in a speech in 2015. He found many factors working on both supply and demand to push prices up.

On the supply side, Canada is a highly urbanized country, and many of our cities have land-use constraints that limit supply, such as green belts and other zoning restrictions. Geography, in the form of mountains and water, also helps to limit supply and support prices.

In terms of demand, several factors have reinforced an extended trend toward higher prices. These include demographics and a long period of low long-term interest rates. But the point I want to stress here is that when you combine a strong desire for homeownership with rising house prices, you will naturally find increasing levels of debt.

Monetary policy, demand and house prices

The connection between low interest rates, rising house prices and increasing debt levels is worth considering in more detail. The goal of our monetary policy is to deliver low and predictable inflation by keeping supply and demand in the economy in balance. If inflation is too low, we can lower our key policy interest rate and expect to stimulate demand for goods and services. When we raise interest rates, we expect to cool demand.

You would expect, then, that relatively low interest rates would lead to strong demand for housing. Looking back, mortgage rates shifted into a lower range in the late 1990s. In part, this reflected a global trend toward lower inflation and interest rates. But it also reflected the fact that the Bank of Canada had built some credibility around its inflation-targeting policy, which began in 1991. Canadians had come to expect that inflation would remain low, and interest rates moved lower accordingly. This is when our long-term rise in household debt took root.

The situation took another dramatic turn in the wake of the global financial crisis in 2008. Central banks slashed interest rates, in some cases to zero and beyond, and kept them at historically low levels for an extended period.

Internationally coordinated fiscal and monetary actions from 2008 to 2010 provided stimulus and helped the world avoid a second Great Depression. But our economy has struggled to gain traction in the last 10 years, not least because our recovery was interrupted by the collapse in oil prices in late 2014. Today, inflation is on target and the economy is operating very close to potential. However, given the lingering effects of the shocks we have faced, the economy still requires stimulus.

Let me make a very basic and important point here. Policy stimulus has a cost, whatever form it takes. Whether delivered by monetary or fiscal policies, stimulus encourages growth by bringing forward household spending and business investment, financed with debt.

I spoke about these debt dynamics in the Purvis lecture two years ago. If fiscal policy takes the lead in stimulating the economy, this can result in a buildup of government debt. If monetary policy takes the lead, this brings about a buildup in household debt. In both cases, stimulus leads to a buildup of debt over time, whether public or private. And excessive debt levels create a vulnerability, making the economy less resilient to future shocks. This is why policy-makers need to consider the debt consequences of the mix of fiscal and monetary policy.


The burden of debt

Ultimately, what matters most is the burden of servicing debt relative to income. In other words, the lower the interest rate, the more debt a given household can afford to carry. For this analysis, we look at the debt-service ratio, which is the required payments of interest and principal expressed as a percentage of income.

Remarkably, the aggregate debt-service ratio on mortgages for Canadian households has been very stable, remaining within a range of 5 to 7 per cent since the early 1990s. What this means is that Canadians have taken advantage of lower interest rates to carry a higher level of debt, thereby keeping the debt-service ratio fairly constant.

You can see how this would arise. Financial institutions are mainly interested in a borrower's ability to service his or her debt out of regular income. So, lower interest rates make it possible to purchase a more expensive home. Further, with improved access to credit-in particular, the widespread use of home equity lines of credit, or HELOCs-it becomes largely a matter for households themselves to choose their overall level of mortgage debt, and to use that debt for a wider range of purposes.

Indeed, Canadians, regardless of their age group, are increasingly relying on mortgages. Among people under 35 years old, the percentage of homeowners with a mortgage has edged higher from about 85 per cent in 1999 to 90 per cent in 2016. For people in the 55 to 64 age bracket, the increase was more dramatic-from 34 per cent to 46 per cent. This casts a new light on that 170 per cent debt-to-income ratio I cited before.

Notice that the 170 per cent figure represents an average across Canadian households. It includes all those who have little or no debt, which means, to make the average level of debt so high, it also must include some very highly indebted Canadians.

In fact, about 8 per cent of indebted households owe 350 per cent or more of their gross income, representing a bit more than 20 per cent of total household debt. These are the people who would be most affected by an increase in interest rates. We are closely watching the vulnerability represented by this group and the debt they carry, and how it poses a risk to both the financial system and the economy. And it is important for these households to understand how personally vulnerable they may be.

In this context, recent changes to mortgage regulations are particularly welcome-including those that require people to show that they can service their debt at higher interest rates. These regulations are helping reduce the economy's vulnerability, since new borrowers will be more resilient than existing borrowers. There are signs that these and other rules are working, as we are already seeing a significant reduction in the issuance of very high loan-to-income mortgages.  

High debt levels and monetary policy

However, these regulations apply only to new mortgages. The stock of household debt, including the $1.5 trillion in existing mortgages, will persist. And this debt has increasing implications for monetary policy. As I said at the beginning, a significant issue for us now is gauging how much more sensitive consumers, and the whole economy, have become to changes in interest rates.

This is particularly important right now because the economy will require higher interest rates over time to meet our inflation goals. Given current levels of household debt, we expect that moves in our policy rate will have a stronger impact in cooling demand than they did in previous years. But this is a significant uncertainty-the sensitivity could be larger or smaller than we expect.

Since last July, the Bank has raised interest rates three times, taking the policy rate from 0.5 per cent to 1.25 per cent. However, it is still too soon to know just how strong an impact these moves will have. There are many reasons why interest rate changes take time-up to two years-to fully work through the economy. For example, consider that the majority of mortgages in Canada have a fixed interest rate, which is usually adjusted only at the end of the term-most often every five years. Those fixed-rate mortgages that have not been renewed since last July have yet to be affected by the interest rate increases. Some of the people renewing in the last few months may have been given a rate similar to the one they received five years ago. Of course, those who have opted for a floating rate-some 25 per cent of mortgages-have already seen their rate resetting higher.

That said, we are seeing some other evidence of the impact of higher interest rates. Banks have increased the interest rates on new loans-not just mortgages, but also other forms of consumer and business borrowing. We have also seen signs that the growth rate of borrowing has begun to moderate.

You may be wondering where interest rates are headed. We know there is some level for our policy rate that is considered neutral-where it will neither stimulate nor cool the economy. This neutral rate cannot be observed, and we do not control it. What is more, it can move around over time as the global and domestic economies evolve.

Despite this uncertainty, it is a useful reference point for central banks, for three reasons. First, the further the policy rate is from the neutral rate, the greater the impact on the economy. Second, because the neutral rate does change, any given policy setting can become less or more stimulative over time, even if the central bank keeps it unchanged. And third, if the neutral rate in an economy falls far enough, it may be difficult for a central bank to provide enough stimulus in the event of a serious downturn.

In our Monetary Policy Report (MPR) last month, we published our latest estimate of Canada's neutral rate, saying it falls in a range between 2.50 and 3.50 per cent, assuming that all shocks affecting the economy have dissipated. At 1.25 per cent, our current policy rate is still well below our estimate of the neutral rate.

With supply and demand in our economy currently close to being balanced, you might expect our policy rate to be much closer to neutral. But several forces appear to be still acting to restrain the economy. We talked about these in the MPR. They include the new mortgage rules, ongoing uncertainty about US trade policy and the renegotiation of the North American Free Trade Agreement, and a range of competitiveness challenges affecting Canadian exporters. These forces will not last forever. As they fade, the need for continued monetary stimulus will also diminish and interest rates will naturally move higher.

Another benchmark for measuring monetary stimulus is the real rate of interest, defined as our policy rate, less the rate of inflation. Today, our inflation-adjusted policy rate stands at -0.75 per cent. As the economy progresses and the forces acting against it fade, the need for an inflation-adjusted policy rate below zero is steadily diminishing.

Managing the risks

All this to say that we are becoming more confident that the economy will need less monetary stimulus over time. Still, as we approach every interest rate decision, we need to consider all the risks the economy is facing relative to our forecast, including those related to household debt.

If we raise rates too quickly, we risk choking off growth and falling short of our inflation target. If we move too slowly, we risk a buildup of inflation pressures that would cause an overshoot of our inflation target. At the same time, moving too slowly would mean a further accumulation of household debt and rising vulnerabilities, while moving too quickly could trigger the sort of financial stability risk we are trying to avoid.

As you can imagine, getting the path of monetary policy right involves a lot of judgment. Bank staff have recently developed an important new way to evaluate these trade-offs and help inform this judgment, and we are publishing a staff analytical note today on this work.

Briefly, the framework uses our models to calculate the risks to the economy associated with various hypothetical interest rate paths. By examining many such paths, we are able to sketch the trade-offs involved in choosing any particular path. Intuitively, higher interest rates will mean slower economic growth; but they will also mean reduced financial vulnerabilities. As a result, the impact on the economy of a major financial stability event would be less.

From this starting point, the framework then allows for the inclusion of macroprodential policies, such as the new mortgage guidelines. By reducing financial vulnerabilities directly, macroprudential policies improve the trade-off policy-makers face in choosing when to adjust interest rates higher. Put another way, macroprudential policies allow monetary policy to deliver similar results for growth and inflation without exacerbating financial vulnerabilities.


It is time for me to conclude. If Shakespeare were writing today, he might say that our financial system gives Canadians more choices than ever in deciding whether to be, or not to be, in debt. Today's record level of household borrowing reflects the evolution of the financial system and the comfort level of Canadians in taking on debt. But it also reflects a prolonged period of very low interest rates and rising house prices.

At the Bank of Canada, we have been watching these debt levels closely because of the growing risks they pose to financial stability and the economy. We know that a portion of Canadian households are carrying large debts, and the concern will become larger for them as interest rates rise. Of course, higher interest rates would likely reflect an economy that is on even more solid ground and less prone to a major economic setback. Furthermore, our financial system is resilient, and the new mortgage rules mean that it is becoming progressively more so. Even so, our economy is at risk should there be an unexpected increase in bond yields or a global slowdown, because both effects would be magnified by their interaction with high household debt.

Ultimately, the Bank's job is to look at the economy as a whole and judge the outlook for inflation. Today, the view is quite good, even with the shadow cast by household debt. This debt still poses risks to the economy and financial stability, and its sheer size means that its risks will be with us for some time. But there is good reason to think that we can continue to manage these risks successfully. The economic progress we have seen makes us more confident that higher interest rates will be warranted over time, although some monetary policy accommodation will still be needed. We will continue to watch how households and the entire economy are reacting to higher interest rates. And we will be cautious in making future adjustments to monetary policy, guided by incoming data.