Front-end Treasury yields flipped from one cut to four hikes since late February

The effective federal funds rate is 3.63% ^[5]. Add four quarter-point hikes and you get 4.63%, which is exactly where the 2-year Treasury closed on Friday ^[7][1]. Wolf Richter writes that buyers and sellers at that end of the market have flipped from pricing one cut in February to four hikes ^[8].

The since-February numbers are what give the front-end story its force. The 2-year is up 126 basis points to 4.63%, so it was near 3.37% in late February ^[7][3]; the 3-year is up 108 to 4.72%, or about 3.64% then ^[4][3][4]. Richter gives those comparisons for the 2- and 3-year but not for the 10- or 30-year, so the move that can be checked across the curve is the week's: 1- to 7-year maturities went 22 to 26 basis points higher in four days ^[1].

Three auctions cleared $119 billion in three days ^[5]. Tuesday's $58 billion of 3-year notes stopped at 4.474% and were 24.6 basis points cheaper by Friday's close ^[2][3][6]. Wednesday's 10-year took 4.834%, the highest auction yield since August 2007, and ended the week 13.6 basis points above that stop ^[9][10][7]. Thursday's 30-year cleared 5.308% and gave up 5.2 more by Friday evening ^[12][13][8].

The mortgage market is where this lands for most balance sheets. At 7.12%, the mortgage rate Richter cites sits 215 basis points above the 10-year at 4.97% ^[14][10][9]. A borrower refinancing in 2026 pays the 7.12%, not the Treasury yield. A treasurer who holds off for the cut is choosing to carry floating exposure at the funds rate plus whatever the front end prices next.

The counter-case has two halves. Richter wrote that "a 10-year yield of 5% is not high, compared to the decades before the Fed's financial repression" ^[17], and in October 2023 the 10-year briefly touched 5% while the funds rate was 5.33% and the curve was inverted ^[21], so the level tells you little about the direction. The cleaner objection is that this price has nothing anchoring it: Richter notes there is no forward guidance now, and that buyers and sellers are steered by the data as they see it ^[20]; he writes that Fed boss Warsh "despises forward guidance and wants the bond market to do its job" ^[19]. Two soft prints and the 2-year could unwind 100 basis points faster than the 30-year, leaving anyone who termed out at 4.72% for three years paying for the insurance. I would still plan on hikes, because the funding cost that matters for 2026 is set at the front end, and the 2-year has already moved 126 basis points since February ^[7].