Warsh swaps the Fed's forward guidance for a test investors have to run themselves

What replaces the path is a conditional. Anyone with a data feed can test it: Warsh's stated standard is that the Fed must be confident underlying inflation is moving to its objective clearly and at sufficient speed, and otherwise, in his words, "we have work to do" ^[15]. That sentence does more work than a telegraphed path, because it tells you the burden of proof now sits on disinflation rather than on the case for holding.

Take the two numbers he chose to read out loud. Twelve-month PCE at 3.7% and the six-month at 4.1% ^[7] sit 1.7 and 2.1 points above the 2% target ^[21], and if the six-month figure is the annualised rate the Fed normally quotes, the earlier half of that same year ran near 3.3%, since (1.037 divided by the square root of 1.041), squared, comes to 1.033 ^[23]. The recent half is the hotter half. He added that progress on broad inflation measures since 2022 has been modest ^[16], and that broad financial conditions are not restrictive ^[13], which is the closest thing to a directional signal in the speech and does not point toward cuts. The 65 months of sustained, elevated inflation he placed squarely with the central bank ^[9] is five years and five months ^[22].

The tape says the street is not paying much for any of it: gold and the VIX each about 1% lower during the speech, longer-dated Treasuries lower too ^[17], and the Dow, S&P 500 and Nasdaq flat in early trading ^[18]. Compare July, when Fortune reports bond yields tracked higher after his press conference as markets digested a Fed on hold and the suggestion that markets were doing the tightening work themselves ^[19]. Yields moved on substance in July; in August, doctrine alone moved nothing.

This is probably wrong, but the retreat looks cheaply priced rather than disbelieved. Withdrawing guidance costs a market almost nothing while everything is on hold and the reaction function is legible; it starts costing when the cycle turns and there is no pre-announced sequence to lean on, and the bill then arrives as term premium and gaps around data releases rather than as volatility on a Friday morning in August.

This could go otherwise in a few ways. Warsh's own framing confines limited guidance to "normal times" ^[4] while conceding the practice was essential during the financial crisis ^[3], which is an exemption he has already drafted for himself; the weakness in the employment outlook for recent graduates that he flagged ^[12] could widen from an aside into the reason he invokes it; or he stays silent on dates while the market simply reads "not restrictive" ^[13] as the telegraph, in which case guidance has been renamed rather than retired.

This is testable, and cleanly so: any pre-committed sequence of moves, or calendar language, in the next few meetings, and the doctrine was posture. Confirmation looks like a decision arriving without a warning shot, with unemployment at 4.1% ^[11] holding well enough that the crisis exemption never gets used. What he is declining to spend, and he said so, is credibility on quasi-commitments that inhibit the Fed's freedom to decide later ^[5]. The discipline is the commitment; the decision itself stays open ^[10], and the practical translation is that the next trade has to be argued from the prints.