The Bank of Canada’s Big Nothingburger

There’s been a lot of commotion as of late over the rapid pace of the Bank of Canada’s current bond buying program, with concern mounting over the speed at which the BoC is scooping up the available supply of Government of Canada bonds. As of December 2020, the Bank of Canada held approximately 37% of outstanding GoC bonds and are currently continuing their Quantitative Easing program at a pace of $4B per week. With Bank of Canada governor Tiff Macklem recently going on record stating that the BoC had capacity to purchase up to 50% of the outstanding supply of bonds before impairing market functionality1, it becomes a relatively simple process to estimate just how much “QE” runway the BoC has left.

Going purely by the quantity of GoC debt outstanding at the end of December, the BoC could theoretically purchase $102B in new government bonds before hitting the BoC’s 50% threshold. According to BMO Economics2, if deficits for the coming year come in on the high side of projections, net bond issuance will be approximately $167B (gross issuance will be $263B to account for $96B of maturing debt). This will allow for BoC bond purchases of roughly $186B ($102B + ($167B / 2)) of new bonds for 2021. At a buying pace of $4B/week, a rough back-of-the-envelope calculation would have the Bank of Canada reaching their self-imposed 50% threshold in approximately 46 weeks. For argument’s sake, let’s just say that at the current pace of QE, the BoC has enough runway to take them to the end of 2021.

What all of this means is that at some point in the next year, if we take Tiff Macklem at face value, the Bank of Canada will need to begin seriously curtailing the rate of bond purchases associated with their QE program. If it’s true that the BoC has been largely responsible for holding down interest rates since the start of the COVID pandemic, then it should be a fairly trivial conclusion to draw that the beginning of QE tapering in Canada should begin to put upwards pressure on long term interest rates.

Of course, all of this assumes that the recent decline in longer term bond yields in Canada are the direct result of the BoC’s quantitative easing program, which certainly makes sense given the Bank of Canada’s stated objective of lowering long term interest rates through QE. For example, rates on the 10-year bond from January 2020 through December 2020 have fallen by about 60bps, and the obvious explanation for this decline is the Bank of Canada’s QE program which has taken the BoC’s ownership of GoC bonds from less than 15% at the beginning of the year to nearly 37% at the end of 2020.

The good folks over at CIBC Economics, however, have a bit of a different take on the recent decline in long term government bond yields3. By decomposing the 2020 change in the 10-year bond yield into the portion caused simply by a declining BoC overnight rate and those caused by broader supply and demand, CIBC deconstructs the change in yield into a “Risk-Neutral” component and a “Term Premium” component. The Term Premium component is the portion impacted by supply and demand issues such as the BoC’s QE program, while the Risk-Neutral component is driven by market expectations of the BoC’s future overnight rate. Of the roughly 60bps fall in the 10-year yield over 2020, CIBC estimates that approximately 15bps are attributed to supply and demand factors (the Term Premium), which would mean that the majority of the drop in yield (45bps), was caused simply by the reduction of the BoC target rate and the markets expectations of a “lower for longer” monetary stance well into 2022 (the Risk-Neutral component).

So, if supply and demand factors contributed to a 15bp fall in the 10-year yield, how much of this was due to the BoC’s QE program? After all, the Government of Canada issued a lot more debt in 2020, so demand for this debt by someone not only prevented rates on the 10-year from rising, but actually caused yields to fall. The obvious answer is that the BoC has been capping rates via QE, but the CIBC report has a different answer for this too. In studying the 2008 financial crisis, a period where Canada did not foray into the world of QE, they note that Canada’s Term Premium effectively followed that of the United States. Said another way, the reduction in the component of Canada’s 10-year bond yield that is caused by supply and demand wasn’t caused by QE in Canada, but was effectively a spill-over of global capital into the Canadian bond market. 

The BoC may not have been undertaking QE to raise bond prices and lower yields in Canada following the 2008 financial crisis, but the fact that the Term Premium component of the 10-year yield declined in line with the reduction in US Term Premiums suggest that many global market players view the Canadian bond market as a reasonable substitute for US bonds and were happy to scoop them up as a result of global QE programs. If the trend from the financial crisis is extrapolated to today, then much of the ceiling being placed on longer term interest rates in Canada isn’t due to anything the BoC is doing, but, rather, is being driven by a global implementation of QE in much larger economies. 

As a small economy with open capital flows, then, the CIBC report suggests that the Bank of Canada doesn’t really control long term yields through QE in as significant a way as many market participants tend to assume. The BoC can certainly impact 10-year yields through its control of very short-term rates as they affect the Risk-Neutral component, but large-scale asset purchases seem to be relatively low-impact. In effect, Canada essentially surrenders much of the control of its long rates to global demand. A corollary of this, if true, is that QE tapering won’t have a serious impact on the 10-year yield and won’t cause rates to spike in its absence – any increase should be small and gradual.

However, this isn’t to say that QE tapering will have no consequences at all. CIBC estimates that had the BoC not engaged in QE, 10-year yields might be about 22 bps higher today than they otherwise would have been. While 22bps isn’t nothing, given how much of the GoC bond market is now owned by the BoC, reducing yields by only 22bps is really as close to a “nothingburger” as you can get. Really, if the BoC intends to continue to step on the 10-year yield, they need to get down on their hands and knees and pray that global QE continues unabated. If they need to depress yields in the absence of that, their best bet is probably to continue to work on the Risk-Neutral component, not the Term Premium. In other words, the most effective way to lower rates on the 10-year is to reduce the short-term overnight rate below its current target rate of 25bps. 

The BoC, of course, has stated on multiple occasions that 25bps is their effective lower bound, but if push comes to shove and they need another 25bp cut, local Canadian QE probably won’t cut it as the BoC is fast approaching their 50% ownership threshold. As such, the BoC will probably be forced to drop their overnight rate closer to zero, despite their current unwillingness to go there. What all of this essentially means is that the sharpest tool in the BoC’s shed of easy money conditions is still its overnight rate. Sitting right on their self-imposed effective lower bound meant QE probably bought the BoC another quarter point cut on longer-term rates. If any more stimulus is needed and Term Premiums fail to decline globally, the BoC will either be forced to go negative or continue to scoop up an increasing share of the bond market – well beyond its stated 50% threshold. The Bank of Canada still has options, but in an ideal world for the BoC, the easiest one is simply outsourcing QE to the US Fed.