Why Goldman Is Fading Last Week’s Bond Selloff

As discussed here last week, the move in US reals seen in and around the long-end drama that accompanied the Fitch downgrade and Treasury’s quarterly refunding announcement argued for a term premium-based explanation for higher yields.

The timing (the selloff accelerated on Thursday) is a bit unsatisfying, but I’ve resigned myself to accepting a “delayed reaction” narrative that folds in Bill Ackman’s long-end short call.

Whatever the case, the move higher in reals was pronounced. Prior to Friday’s NFP-inspired rally, 30-year reals were the highest in a dozen years.

“[The] US yield curve bear steepen[ed] on higher real risk premia despite [a] lack of fundamental news,” Goldman’s Praveen Korapaty wrote, in his weekly, adding that the move had “the hallmark of bond risk premium repricing,” a characterization consistent with one of the bank’s models.

Note Korapaty’s skepticism around the notion that the market actually learned anything new over the last several sessions. I expressed similar sentiments on multiple occasions throughout the week.

He went on to enumerate each factor cited for last week’s bond selloff (again, pre-payrolls), starting with the idea that some of it was attributable to a ripple effect from the BoJ’s decision to widen the YCC band.

Goldman’s estimates for spillover from JGBs fall short of explaining the move in US bonds even in a scenario where JGB yields rose another 15bps. Further, Korapaty pointed to disaggregated timezone data which shows the bulk of the selloff occurred during New York morning trading. That’s shown in the figure on the right, below.

It’s possible, he conceded, that “other investors are repositioning in advance of anticipated selling from Japanese investors,” but if that’s accurate, then the move isn’t sustainable unless Japanese selling actually materializes in a meaningful way, which Goldman doubts.

Moving along, Korapaty addressed the notion that markets are simply repricing recession odds (lower) via higher bond yields. If that were the case, though, you’d expect more movement at the front end. “Deep [Fed] cuts remain priced, suggesting that market-implied recession probabilities have not changed that much,” he remarked.

The figure below illustrates that latter point. As far as the everyday investor’s concerned, there’s been no real repricing of the Fed trajectory since late last month.

Looking out to the back half of 2024 and 2025, pricing firmed up a bit, but not enough to suggest a material rethink consistent with the outsized jump in long-end bond yields.

Relatedly, some suggest markets are now coming around to the idea that growth in the US won’t just hold up, but may actually continue to run in line with potential, particularly if the fiscal backdrop doesn’t turn more forbidding. Under such macro conditions, bonds could see a “reflationary pricing and steeper curves,” Korapaty remarked, before casting doubt on that notion. “Here too, we think there is inconsistency — after all, if the economy is indeed very strong, we should not expect the Fed to ease quite as much, which would mean the front end needs to reprice higher as well,” he wrote.

Finally, he addressed the supply argument. “While supply was indeed large, it was actually less tilted towards longer maturities, and issuance in duration terms was not too far off expectations,” Korapaty said, of the refunding announcement.

All of that to make a simple point. While Goldman generally agrees with a structural argument around higher medium-/long-term forward yields, higher risk premium, and steeper curves, they’d fade the recent moves. As noted above, the bank doesn’t buy the narrative, and anyway expects “significant cooling” for both inflation and growth going forward, which Korapaty noted “is typically not conducive to a sustainable reset higher in yields.”