Fortune column pins the worst Treasury month in four years on Warsh's hawkish Fed

The column's strongest evidence is the calendar. It calls inflation, deficits and geopolitics background risks and says nothing in August or September re-priced them ^[17], so the trigger has to be something that did move in those weeks. Torsten Slok, Apollo's chief economist, wrote after Jackson Hole that the Fed "went into 2026 expecting several cuts, and now the FOMC is leaning toward hiking" ^[5]. At the September meeting Warsh promised that "this Fed will deliver price stability" ^[8].

The column's bank estimate points the same way. Of the $180 billion in added unrealized losses it models for the third quarter, $115 billion fell in September ^[19]. September alone accounts for about 64% of the quarter's modeled damage, in the first full month after the speech ^[24]. The model is the author's own, built with help from Claude, the column says ^[19].

The Volcker comparison claims more than the pricing does. Markets priced an 80% chance of at least 100 basis points of hikes beyond the path expected before Jackson Hole ^[9]. The 2022-23 cycle was 525 basis points in 17 months ^[10]. The extra tightening priced is about a fifth of the last cycle ^[14]. That is a large repricing for a few weeks and a modest one next to the hikes traders are reacting to. The column itself concedes that 2026 is not a replay of 2022-23 and that the market's fears may be overdone ^[22].

The term premium is where the column argues against its own evidence. It describes the term premium as the extra yield needed to offset duration risk for holders of longer bonds ^[18]. Then it explains the panic by saying traders burned once have good reason to shed long-duration risk ^[23], and the burn it means is 2022-23, when the 10-year lost 16%, its worst return in a century ^[10]. Selling long bonds until their yield pays for the duration risk is a rising term premium. We think the Fed is the cause and the term premium is the route by which a policy repricing reached the 30-year, so the choice between them is partly false. The basis trade is a separate question. The column calls the change there a "modest slackening" and does not say how large it was ^[16].

If the Fed path is what moved yields, long bonds should recover when the priced hikes fail to arrive. Traders demanding a permanently higher premium would keep the 30-year up even after the Fed stops hiking. And if traders misread the August CPI report, as the column allows they may have ^[6], part of the move unwinds with later data. We think the Fed reading is the strongest of the three, because the timing is tight and the column found no other driver that changed in those two months ^[17].

The test that would prove it wrong is the shape of the curve. A policy repricing should lift two-year yields more than 30-year ones, and the column does not cite short-dated yields. If the 30-year rose faster than the two-year in September, something besides the Fed path is pushing up the long end.