Fed minutes cast September's quarter-point hike as insurance against sticky inflation

The middle of the new 3.75%-4% range is 3.875% ^[1]. Subtract August's core PCE estimate of 3.4% and the policy rate is about 0.475 points above underlying inflation ^[21]. Against the headline 3.8% estimate the margin is 0.075 points ^[22]. These are rough, backward-looking real rates. They belong to a committee that judged inflation risks skewed to the upside ^[5], and in July its minutes recorded some officials warning that financial conditions might not be tight enough to bring inflation back to 2% ^[19]. The one further increase most participants see room for ^[2] would lift the core-adjusted margin to about 0.725 points ^[23].

It was the Fed's first increase in three years and two months, Seoul Economic Daily reported ^[9]. The minutes stated that "many participants emphasized that a higher path for the target range would be prudent on risk-management grounds, providing insurance against inflation remaining persistently above target due to stronger-than-expected demand or further adverse supply shocks" ^[4]. Insurance implies cover that can lapse once the risk passes, and the minutes keep that option open: no path is predetermined, and officials will weigh fresh information at each meeting ^[3]. The pressures they named were energy prices tied to geopolitical developments and surging AI-related investment ^[11].

The rest of the year splits three ways. Inflation could cool enough that the remaining increase never comes. Steady data would put it at the December 9 decision ^[10]. The third case breaks the insurance framing: the upside risks that several participants said had grown more pronounced in recent months ^[7] persist until the officials who argued from the baseline outlook ^[6] win the argument, and one more hike becomes several. I think the second case is the likeliest, mostly because the futures market already puts it there. Post-minutes pricing leaves about a 17% chance of any October move ^[20], and investors expect the Fed to avoid raising rates before the Nov. 3 midterm elections, according to Seoul Economic Daily ^[13]. The election falls between the October 27-28 and December 9 decisions ^[10].

Officials have said much the same in public. New York Fed President John Williams said on Sept. 29 that one more increase by year end could be appropriate but that there was no need to rush, Seoul Economic Daily reported ^[14]. The labor market lets the committee wait. Participants judged it close to maximum employment, with unemployment at 4.1% in July and August and the risks to employment balanced ^[17].

The committee is also not planning for 2027. Of the 18 officials who submitted projections, 16 expect one more hike this year and none next year, according to CNBC figures cited by Quartz ^[12], or about 89% of the forecasts ^[24]. Chairman Kevin Warsh, who joined the Fed in May, is not among the 18 ^[15].

The insurance reading is wrong if the Fed moves on October 27-28 against those odds ^[10], or if the officials now projecting no 2027 increases start projecting one ^[12].