The policy rate’s in the neighborhood of neutral and having delivered 75bps of cuts since September, the Fed’s in a position to wait and see how the US economy evolves.

That was the overarching takeaway from Jerome Powell’s press conference on Wednesday, when a divided Fed lowered rates by a quarter-point for the third straight meeting.

Analysts, financial commentators and, of course, AI models, were at pains to parse the Fed communications for any and all nuance, a quixotic effort that produced a quasi-consensus centered on the notion that somehow, i) two hawkish dissents, ii) an unchanged projection for just one cut in 2026, iii) a half-point upgrade to the near-term growth outlook and iv) a 12-month UNR projection of 4.4%, constituted a dovish spin on Wednesday’s move.

If you ask me, the fraught nature of the private debate (in all, half a dozen participants in this week’s proceedings favored no change in rates) and the public communications netted to a hawkish cut, consistent with market expectations headed in. When you consider how hard the Fed works to align market expectations with its intentions for each meeting, that interpretation seems like the best one to me.

Plainly, T-bill purchases to alleviate funding market strains and prevent passive tightening through the reserves channel, don’t translate directly into a dovish policy decision. To the extent any AI models might’ve picked up on the reserve management headlines in the course of “scoring” this week’s proceedings as dovish, methinks the models need tweaking.

Powell confused the issue further during the press conference when he carried on about phantom jobs growth. If, as he postulated, payroll gains are negative and anyway overstated by ~60,000, then why not take the hint from the deluge of private-sector data (which all points to a weakening labor market) and try to come up with a set of communications that doesn’t suggest the jobless rate will stay glued to what, historically, counts as full employment while the economy expands at a 2.3% clip?

Anyway, who cares, right? None of this is going to matter in six months, let alone 12. Hell, it may not matter in six days, when the Fed will learn what the BLS reckons for headline hiring in November. In the same release, the government will offer a partial jobs report for October.

But Powell said Wednesday the Fed isn’t inclined to trust the data for some of the same reasons no one else trusts it — namely, it’s distorted by the shutdown. He said the same of the next CPI release and indicated it’ll be at least a month (or two) before anyone has a clean read on… well, on anything.

In the meantime, the Fed’s working assumption is that between a resilient high-end consumer and AI spending, the US economy’s probably fine. That’s what Powell told Reuters’s Howard Schneider during the presser. “Fiscal policy’s going to be supportive” early next year, he added, during the same exchange.

Asked by Steve Liesman whether the Fed’s taken out enough insurance over a trio of so-called “risk management” cuts since September, Powell said “yes.” “We’re well-positioned to wait and see how the economy evolves” in lieu of missing government data.

When Colby Smith asked if dissents indicate a higher bar for rate cuts going forward, Powell launched into a recap of what it means when the Fed’s goals are in tension. “Everyone agrees inflation is too high” and everyone agrees that risks to the labor market have grown, he said. The difference is in how participants weigh those risks.

He went on to suggest that nine out of 12 ain’t bad when it comes to marshaling support for a given policy decision under the circumstances.

Smith pressed on, wondering if dissents become “counterproductive” at some point. Powell said no. “We always hope the data will give us a clear read, but we have a competing” mandate, he told Smith. “You’ve got one tool. You can’t do two things at once.”

Here’s hoping Powell didn’t pop a stick of Winterfresh before he walked out of the building.