Last year was bad for bonds.
In fact, we can call 2022 anomalously poor. If you use datasets which splice together historical series to create the rough equivalent of today’s 10-year Treasury, you can conjure a visual where 2022 sticks out as among the worst years in the history of the republic for US government debt.
A third consecutive annual loss would be an unprecedented event (see the figure below), but as you might’ve heard, Bill Dudley expects more pain. “I have what might be disconcerting news: It’s not over,” he wrote, in a Bloomberg Op-Ed that’s garnering more attention that it otherwise might thanks to a rebuttal from Morgan Stanley’s rates team.
To be clear: Precision forecasting is even more impossible for Treasury yields than it is for stock benchmarks, assuming there are degrees of “impossible.” Keep that in mind as you read further.
Morgan Stanley’s Matthew Hornbach, along with colleagues including Guneet Dhingra, likened Dudley’s bond call to a dire prognostication from Jeremy Siegel a dozen years ago. “Dudley’s prognosis reminds us of a similarly bearish view espoused by Professor Siegel in 2010,” Hornbach wrote, referring to Dudley as a “pundit” along with the obligatory reference to his storied career.
In 2010, Siegel warned that bond bulls were “courting disaster.” That call turned out about like every other bearish bond call looking back four decades: Poorly.
Notwithstanding the impossibility of deriving accurate year-end point forecasts, you could’ve done a pretty good job projecting the general direction of bond yields over time by simply reading a Hoisington quarterly and extrapolating it into the indefinite future. (Right up until in 2021, that is.)
Dudley cited three factors that he said should bias yields higher. One is the Fed’s avowed inclination to keep hiking rates or at least to hold terminal for an extended period of time. He mentioned the dots, and specifically the fact that seven participants now have a higher long run dot. “They appear to be increasing their estimate of the neutral rate [which] makes sense,” Dudley wrote, flagging spending by boomers, government deficits and the likelihood that “vast” sums will be invested in securing supply chains and facilitating the green transition.
He went on to say inflation will “almost certainly” be higher than 2% going forward and suggested the term premium should probably be higher given uncertainty around inflation and fiscal policy. Ultimately, Dudley concluded that a “conservative” estimate for 10-year yields is 4.5%.
Spoiler alert: Forecasting bond yields is harder than simply adding up three components as though they’re independent of one another and semi-static. Dudley knows that, of course, but Bloomberg Opinion isn’t exactly the forum for 50-page academic papers. Suffice to say Dudley’s short column was begging to be picked apart by some rates team somewhere, and Morgan Stanley couldn’t resist.
“The last time the real funds rate averaged 1% over 10 years was the decade ending September 2010. Most of that 10-year window occurred before the financial crisis [and] that 10-year period saw nonfarm productivity rise 2.7% annually,” Hornbach wrote, addressing the neutral rate component. Morgan Stanley conceded it’s possible that real Fed funds will average 1% over the next 10 years, but they doubt it. They’re sticking with 0-0.5%.
As for the notion that inflation will average, say, 2.5% over the next decade, Hornbach doubts that too. CPI annualized at 2.7% over the last 10 years, but that includes an ostensibly extraordinary period (“ostensibly” because it could be that what we’re experiencing now is “normal” and the Great Moderation was actually extraordinary).
Morgan Stanley doesn’t expect hot nominal growth, super-charged fiscal policy and “runaway inflation” (Hornbach’s description) to last, or at least not for a decade.
As for the term premium component, Dudley’s guess is as good as Hornbach’s is as good as Dudley’s is as good as Hornbach’s… you get the idea. This is a notoriously difficult debate, and the QE era complicated it immeasurably.
Hornbach noted that even as G4 central bank balance sheets are set to keep contracting, they’ll likely still be a third larger compared to 2018 by the end of next year. In light of that, the bank “see[s] very little scope for term premiums to rise,” and besides, there’s something inconsistent about higher short-end rates and rising term premiums.
Ultimately, Morgan Stanley expects the long-term range for 10-year yields to be between 2-3%.
Guess what? I’m inclined to side with Dudley. That won’t come as a surprise to a lot of regular readers. I think economists and analysts are too wedded to the notion that the Great Moderation was a “new normal” rather than an anomalous period of macro stability enabled by structural factors some of which (but not all) were upended by the pandemic and the war.
Simply suggesting that inflation is going to return to 2% and bond yields will revert to “normal” levels in a changed world seems a little too convenient for me. The neutral rate is probably higher (the performance of the US economy in 2023 strongly suggests as much), term premiums probably need to rise for precisely the reasons Dudley suggested and inflation likely isn’t going back to 2% on a sustainable basis. At the least, I’d expect inflation volatility to remain elevated in perpetuity.
Importantly, that doesn’t mean Dudley’s “forecast” for 10-year yields will be right or that Morgan Stanley’s will be wrong. Dudley would’ve been (far) better off making the case without the precise yield projection. His argument is a good one, but a good way to turn a good argument bad in the market context is to attach a precision forecast to it. As Hornbach himself put it in 2018, “history has shown consensus estimates for Treasury yields are usually wrong [and] everyone understands that accurate point forecasts rarely occur.”
