Odds may now favor a “sharp reversal” in US long-end yields.

That’s according to Goldman, but it’s also according to common sense. Even if you think the worst bond selloff in at least a generation (and quite possibly in American history, depending on what, exactly, you mean by “US government bonds”) has further to go, or that the fundamentals argue strongly for structurally higher yields in perpetuity, extreme market moves over compressed time frames are typically good fades. The problem is picking the local top (or bottom).

As regular readers will attest, I’m fully on board with most of the arguments for structurally higher yields, I think the term premium repricing was entirely rational and I think r-star is probably higher. My pushback is on the idea that it’ll be a one-way ticket forever — that no meaningful bullish correction is coming even in the event the US data turns. That seems implausible to me, or at least unlikely, which is why I’ve carefully suggested duration could find and sustain a bid in the not-so-distant future.

For their part, Goldman reiterated that yields are above fair value, even more so after last week’s selloff. To wit, from the bank’s Praveen Korapaty:

10-year USTs now sit meaningfully above our fair value estimate of 4.2-4.3%. On fundamental grounds, we see risks for rates as tilted to the downside from here. Current medium-term real yields are already above the long-run potential growth rate and our expectation of a ‘Q4 growth pothole,’ which should become more apparent in the data towards year-end, we noted that yields tend to partially reverse after large selloffs, and using the average historical pattern would suggest 10-year yields top out at around 4.6-4.8%, and could retrace back to around 4.1-4.2%.

Again, I really don’t think there’s a lot to argue with there. Expecting a rally from these levels doesn’t require any sort of outright bullishness. Bonds are just an oversold asset class, and all falling knife jokes aside, I think the odds of a reversal are only going to rise if the selloff does manage to extend.

Korapaty conceded that historical experience, fair value estimates and key levels are probably “imprecise guideposts at best” in the current environment. “A more reliable way to gauge when yields may be locally topping out is to look for potential catalysts — a stretch of soft economic data, rising pressures in other growth-sensitive financial assets (like equities), or even rising financial stability concerns,” he wrote.

Goldman pushed back on the notion that the selloff is disorderly, and rather suggested it’s more buyer’s strike than anything else. “Yield curve dispersion and liquidity metrics have not deteriorated noticeably” and although the (by now well socialized) negative convexity around CTD switches may be leading to higher volumes in the 10- to 20-year sector, Korapaty said “the actual DV01 delivered to investors owning these futures contracts because of this extension isn’t large enough to have played a major role in the selloff” in Goldman’s view.

As for stop-outs, or overwhelming supply in the face of insufficient demand, Goldman doesn’t really see that either. “Ordinarily, such imbalances [would] clog up dealer balance sheets and lead to a deterioration in liquidity metrics, and stop-outs show up in elevated volume data,” Korapaty went on.

So, what’s the problem? Well, from where Goldman’s sitting, “the rapid re-adjustment of price appears to be allowing for an orderly clearing of markets and to be compensating for lagging buying interest.”