Color me amazed (and grateful as a long-only investor) that US equities continue to hold up in the face of what’s on track to be another very bad month for bonds.
As of Thursday, the S&P’s more or less flat for September, a month during which 10-year US yields are up 40bps and (still) counting.
To be sure, the bond selloff probably is inhibiting stocks (i.e., capping index upside), but it hasn’t derailed them. And that’s saying something, particularly given the back-half of September tends to be the worst two-week stretch of the year for the S&P, as corporates enter the buyback blackout ahead of Q3 reporting season.
The figure below gives you some context for the surge in 10-year reals.
The rolling one-month increase for reals is among the steepest of the last two and a half years, and yet the S&P’s just 1.3% from its record high.
You can chalk that up to a number of things, not least of which is the idea that equities are actually “cheap” — or at least not expensive — on a forward multiple. That logic holds right up until profit expectations roll over, at which point the index would re-rate mechanically. That’s the “Fata Morgana” risk discussed here.
Another excuse says bonds are selling off for the “right” reasons — that despite appearances, yields aren’t rising in a “disorderly” fashion, but rather responding to incremental evidence that the US economy’s still firing on most cylinders and that the labor market’s generally fine.
Note that jobless claims printed sub-200,000 for a second week on Thursday, while ADP’s weekly update on private sector hiring showed employers added an average of 20,000 jobs per week in the four-week period to September 5 (the most recent observation), the quickest pace since June.
At the same time — and Morgan Stanley’s Mike Wilson mentions this on a regular basis — the term premium isn’t evidencing anything like panic about America’s fiscal trajectory.
In fact, at ~65bps on the latest update from the New York Fed, it’s 20bps below this year’s high, as shown below.
As the figure reminds you, we’ve been in positive territory for the duration (no pun intended) of Trump’s second term, but there’s nothing inherently worrying about that.
Indeed, you could argue 65bps is a relative pittance in terms of what investors “should” be demanding under the current circumstances to loan the US money for 10 years versus just rolling short-term Treasury paper.
If you’re surprised to learn that the term premium’s lower despite the escalatory rise in long-end yields, it’s worth noting that the ACM model (shown above) isn’t the only estimate. Another estimate — which, in addition to information from the curve, incorporates data on blue-chip forecasts for future short-term rates — puts the term premium at 96bps.
It’s tempting to call that the more accurate guesstimate given what we think “should” be the case considering debt, deficits, political dysfunction and so forth. But even there, 1%’s not historically unusual — it’s only “high” in the post-GFC/QE context.
All of that to say it’s possible to make the case that the bond selloff isn’t as worrying as it most assuredly looks and feels, and that stocks somehow “know” that.
Regardless, there’s only so much of this (i.e., higher yields) equities can take without at some point selling off to reflect the mathematical impact of a sharply higher discount rate. Or so textbooks would tell us.
Meanwhile, Thursday’s seven year auction was weak, with the lowest indirects since late last year.
Thursday’s seven-year offering was at least better received than the prior day’s abysmal five-year sale, but it still tailed despite being the highest-yielding seven-year auction since the note was brought back early in 2009. As BMO’s Vail Hartman noted, the WI rate was more than 50bps above August’s stop.
So — and as ever — the truth’s probably somewhere in-between. The bond selloff isn’t irrational, nor necessarily out of control. And yields are rising for the “right” reasons, some of which are bullish (or at least not bearish) for equities.
But past a certain point, higher yields will become a valuation headwind, stocks’ purported “cheapness” on a forward multiple notwithstanding.
It is everything all at once. Rates are resetting due to inflation and an abundance of higher yield offerings from the hyper-scalers and others scrambling to fund the AI buildout. It is the war(s), tariffs, and a loss of confidence in our government. It is the bond vigilantes turning their nose at certain offerings, and it is the price of oil (and diesel) dramatically altering everyone’s supply chain calculus every time there is another ME escalation in the news.
I think we are certainly due for a correction, as some of the challenges confronting the AI build-out become more apparent. AI (the main linch pin still holding everything together) cannot continue to grow exponentially if the data centers they require cannot be built on time. Political backlash will eventually lead to state and local regulations, law suits, and very costly delays, as the construction costs will only rise over time. Additionally, the available capital to fund those things is becoming an issue. Massive issuance and declining bond ratings means higher coupons. At some point, unfinished projects will struggle for funding, and newer projects will have to be delayed (perhaps permanently).