Powell Risks Prejudgment Error With Flippant November Promise

Now that Jerome Powell has (wisely) taken out a 50bps insurance policy against a softening US labor market, it’s time to ask why, in the absence of additional evidence to suggest the labor market is in fact softening, the Fed should be in a hurry to cut rates again in November, let alone by another 50bps.

Powell said repeatedly on Wednesday afternoon that the Fed’s “not in a rush” and that the dot plot doesn’t suggest they are. But notwithstanding the fact that two “dots” (i.e., people) see no additional cuts this year, and seven only one more cut, the “median” official apparently is in a hurry. Because 100bps in three months is a lot of easing to telegraph when current-quarter growth is tracking 3%.

JonesTrading’s Mike O’Rourke captured it well in a characteristically unsparing review:

While repeatedly making comments about the strength of the economy, Powell was asked specifically “What do you expect to learn between now and November that will help inform the scale of the cut at the next meeting?” Common sense would tell most “data dependent” people to respond to such a question by saying a cut may not be needed, before delving into details. Instead, Powell responded “You know, more data, the usual. Don’t look for anything else…” The carefree response was as if the November rate cut is a done deal.

To be fair, I don’t think Powell meant to suggest November’s a done deal for a cut. Outside of the pandemic (when the Fed was committed to keeping rates low) and the 2022-2023 hiking cycle (when keeping longer run inflation expectations a semblance of anchored demanded the Committee be clear about its intention to ratchet rates steadily higher, and as quickly as possible), Powell’s generally careful not to prejudge future meetings. That doesn’t mean he hasn’t accidentally done just that (God knows he has), it’s just to say Powell’s not a Chair who’s inclined to purposefully pre-announce decisions that haven’t been made.

If you listen to the exchange O’Rourke referenced, Powell was just insufficiently attentive to the question, in my view. I think he might’ve missed — or glossed over — the “…at the next meeting” part. I don’t think he meant to pre-announce at least a 25bps cut for November. Rather, he was probably just saying that data will inform the scale of any and all future cuts, and that a hypothetical (likely, but still hypothetical) November cut wouldn’t be any different in that regard.

Of course, an error of omission is an error all the same. So, O’Rourke’s exactly right: Powell should’ve listened closely to the question and presaged his response as follows: “First, let me just be clear: No decision has been made about the next meeting, nor about any other future meetings…”

With all of that in mind, initial US jobless claims dropped 12,000 in the week to September 14, according to Thursday’s release. 219,000 was the lowest since May 18 and it undershot every economist estimate.

The four-week average, at 227,500, is now the lowest since early June. As a reminder: Claims would need to be north of 300,000 to tip any sort of real problem, assuming past is precedent (admittedly not a safe assumption).

Do note: This release covered NFP survey week for September and as such “holds greater relevance,” as BMO’s US rates team put it, adding that the moving average is “comfortably within the pre-pandemic range and offer[s] little reason to assume the labor market is on precipice of a sharper deterioration.”

Right. Continuing claims likewise undershot at 1.829 million for the week to September 7. That was also a four-month low.

In Wednesday’s Daily, I gave Powell a lot of credit. And I think he deserves it. But he needs to be careful here. Defeat can still be snatched from the jaws of victory.

Make no mistake: 50bps was the right move this week. Unequivocally. I was a staunch advocate of a half-point first cut. But it’s far — far — less clear that following up with another cut in less than two months is the best course of action.

The problem, as O’Rourke correctly identified, isn’t that 10 Fed officials suspect it might be appropriate or necessary to cut rates by another 50bps in 2024. The (potential) problem is that effectively promising a cumulative 100bps of rate cuts over just 90 days (that’s more than 1bps per day) when there’s no definitive evidence to suggest the economy needs anything more than the “recalibration” Powell delivered this week, could be a policy mistake in the making.

Critically: Such a mistake would be an unforced error. As it stands right now, on September 19, there’s no obvious reason to rush into another cut. If Powell wants to argue November 7 doesn’t count as “rushing,” I suppose I’d ask what does? That’s the next meeting.

No, the US economy doesn’t need overtly restrictive policy anymore. But given the Fed just cut 50bps — checks watch — yesterday, and given extreme ambiguity around the location of short-term neutral, the Fed shouldn’t effectively promise cuts at every meeting on the way back down the mountain. Or at least not in the absence of clear evidence that the labor market’s tanking.

As Andrew Bailey put it Thursday, when the Bank of England voted 8-1 to keep rates on hold after their own first cut of the cycle, “We need to be careful not to cut too fast or by too much.”

If the Fed’s not, as Powell variously insisted, “in a rush,” officials need to make that clear over the next few weeks during what I’m sure will be dozens of public speaking engagements and business television cameos.