When the Bank of Canada eventually gets around to raising its benchmark interest rate, how quickly will other borrowing costs follow? According to new research, that could depend on the mood of global investors.
Bank of Canada economists Gregory Bauer and Antonio Diez de los Rios have developed a new model for evaluating the extra yield investors demand to hold longer-term bonds. This "risk premium" tends to reflect the economic cycle: when the economy is strong, investors are less inclined to bid up yields; when the economy turns bad, bond buyers require sweeteners to compensate for the risk surrendering their money for a fixed term with no way to get it back.
Understanding how markets price risk is crucial for policy makers. With conventional policy, central banks have only direct control over short-term rates. However, it's assumed that a shift in the cost of borrowing for, say, one year will ripple through the system, raising the cost of borrowing for longer periods.
In the years ahead of the financial crisis, this relationship broke down. Even as the Federal Reserve lifted its benchmark rate in 2004, the yield on 10-year Treasuries fell. At the time, the Fed was stumped. (Alan Greenspan, the Fed chairman, famously called it a "conundrum.")
Later, researchers determined that the reason longer term borrowing costs were falling even as shorter term rates were rising was because buoyant global investors had stopped asking for a risk premium.
The yield on 10-year, zero-coupon Treasuries in May, 2004, was 4.74 per cent. According to Messrs. Bauer and Diez de los Rios, investors' expectations for future interest rates equated to a yield of 4.05 per cent. The extra is explained by a 0.69 per cent risk premium. A little more than a year later, the yield on one-year debt had increased more than two percentage points, yet the yield on the 10-year security had dropped to 4.33 per cent. The reason: the risk premium disappeared, dropping to -0.02 per cent, according to the Bank of Canada economists.
The work of Messrs. Bauer and Diez de los Rios confirms previous findings about U.S. rates. However, they expanded the analysis to test whether the risk premium has a similar effect on longer term rates in Canada, Britain and Germany. Their model suggests that it does. They also showed that the relationship exists when investors are fearful by comparing one-year and 10-year rates in each of the last five U.S. recessions. Between October 2007 and June 2009, for example, the yield on Canada's 10-year debt fell less than 0.5 per cent even as the Bank of Canada reduced shorter term rates by more than three percentage points. According to Messrs. Bauer and Diez de los Rios, the risk premium rose more than two percentage points over that period, mostly offsetting the drop in interest-rate expectations.
Perhaps the most important thing Mr. Bauer and Mr. Diez de los Rios have done is provide empirical evidence that the mood of global investors has a significant impact on longer-term interest rates in individual countries.
Another point that comes through in the paper is that these investors will think for themselves: if they are wary about the prospects for the global economy, they won't be swayed easily by central banks cutting short-term rates.
That helps explain why the Fed has had to resort to buying bonds itself to lower longer term rates. And it could explain why Bank of Canada Governor Mark Carney continues to talk about raising interest rates, despite a troubled global economy.
International investors see Canadian debt as a haven, demand that is putting downward pressure on longer term interest rates. Under those conditions, the Bank of Canada might have to raise the benchmark rate considerably in order to raise overall borrowing costs enough to stamp out inflation. That would argue for raising the overnight target from its current 1 per cent sooner rather than later, since global demand for Canadian debt could keep borrowing costs low for some time yet.