September 22, 2023 | 13:00
It’s Fall, Except for Yields
Ah, September. Crisp nights, cool weather, pennant races, NFL football, and rocky equity markets. Fully living up to its reputation as the most challenging page on the calendar, this month has already seen a roughly 4% pullback in the S&P 500 as of Thursday, putting it on track to be the worst month since… last September, when it fell 9.3%. (Technically, Dec/22 saw a 5.9% drop, but let’s not fuss with a good story.) There is no mystery whatsoever behind the latest market angst—bond yields ratcheted higher again this week, with the 10-year Treasury yield cracking the 4.5% threshold for the first time since 2007. While Friday saw some relief, yields are still up massively in a matter of months—recall 10s were below 4% as recently as the end of July, and 3.5% in mid-May.
The forced upward march in yields has been driven by the dawning realization that U.S. growth is amazingly resilient, underlying inflation is not going quietly into the night, and oil prices are still a nagging concern at around $90. This week’s trigger for another upward lurch was the FOMC, even though a hawkish pause was universally anticipated. The Fed managed to out-hawk expectations with fewer expected rate cuts in 2024 in the dot plot. While we all know better than to accept the dot plot as gospel, it nevertheless served as a fresh excuse to sell bonds.
Canada’s bond market had an even more challenging week, with yields jumping nearly 20 bps across the curve. Like Treasuries, a variety of maturities hit levels not seen since the days before the Global Financial Crisis blew open in 2007—5s at 4.25%, 10s near 4%, and 30s at 3.75%. Just to keep some perspective, 10-year GoC yields never rose above 2% in the year before the pandemic, and plumbed lows of less than 0.5% in 2020 (Chart 1).
Aggravating the spillover from the U.S. selloff, the domestic market also grappled with a high-side surprise in August CPI. The spike in headline inflation to 4% wasn’t a huge shock, coming in just a tick above our call, but more unsettling was the upswing in core trends. Among the many metrics, perhaps the cleanest message was from the pick-up in median prices to 4.1%; simply, more than 50% of the basket is still seeing inflation of more than 4%—not good. Suffice it to say that the ugly CPI result landed with a thud on a heavy market, as the market ramped up the possibility of one more hike from the BoC later this year.
A speech from Deputy Governor Kozicki sounded a calming note on the very same day as CPI, even if the market focused on any hawkish aspects. She helpfully noted that some volatility in CPI was “not uncommon”, that past rate hikes “will continue to weigh” on activity, and that signs of strain in consumer finances are beginning to emerge. However, the speech also concluded that the Bank was fully prepared to hike rates again, if needed, a message that was echoed in the meeting deliberations. While not new news, that was the major takeaway for markets, and the TSX was drilled with a 4% setback for the week as yields flared higher.
Prior to this week’s events, we had been of the view that the Bank of Canada and the Fed were likely done raising rates, albeit with significant odds of one final 25 bp hike. And, if anything, it seemed like a greater certainty that the Bank was done, what with GDP falling in Q2, still soft in Q3, and the job market more clearly losing steam. Even with higher headline inflation, for now, Canada’s core inflation remains south of U.S. trends, at least on CPI ex food & energy (3.6% y/y vs 4.4%). Moreover, the Canadian consumer faces much more immediate pressure from past hikes due to a higher debt burden and a shorter duration on that debt. Yet it’s notable that even if retail sales match the flash estimate for August (of a 0.3% drop), they will still be up 1.5% from a year ago, while U.S. sales were up 2.5% y/y last month—softer yes, but not a massive gap.
The bottom line is that the risk of further rate hikes has risen for both central banks, but we are not yet at the point of adjusting our call. There’s plenty more economic data to go before the next meetings, and new risks are emerging for the growth outlook in both economies. While Canada looks to avoid an auto strike, the industry will still be swept up by the UAW’s broadening action. Geopolitical risks also keep intruding, with friction between India and Canada the latest flashpoint (for the record, India accounts for just under 1% of the trade flows with Canada, and roughly 40% of international students). Overall, we believe short-term rates are now restrictive enough to do the job, and the pronounced back-up in long-term yields adds another layer of tightening.
Studying the Fed’s dot plot is the modern-day equivalent of Kremlinology, but probably a little less worthy of one’s time. Still, there are always curiosities, and one dot in particular stood out to these eyes—for the end of 2024, one lonely FOMC participant has pencilled in a higher Fed funds rate than for the end of this year. Forget about rate cuts, this member—we suspect Governor Bowman, though we may never know for sure—thinks that rates need to rise another 75 bps from current levels, taking the mid-point to 6.125%. That may seem wild, but consider this: At the FOMC meeting one year ago, not a single member pencilled in rates above 4.875% for end-23. So, not one person was within 50 bps of today’s actual level, not even some of the loudest hawks.
This is certainly not meant to poke fun or question anyone’s forecasting ability; see “people in glass houses”. It’s more to drive home the point that no one saw rates moving this high, even by late last year, when we already had a taste of 9% inflation, $120 oil, and 75 bp Fed rate hikes. The takeaway is that we all need to keep an open mind on where rates are headed in the next year, given the wide range of possible outcomes for both growth and inflation. The consensus is shifting again as we speak to high-for-longer but is still looking for rates to dip 50-to-75 bps from current levels, which jibes with our view. The experience of the past year tells us that this perfectly reasonable view may be perfectly wrong—that lonely 2024 dot may yet prove prescient.