Bond buyers put the 3-year 95 basis points above the funds rate the Fed just raised

The two front-end spreads do not quite agree with each other. A 3-year at 4.86% sitting 95 basis points above the effective federal funds rate implies a funds rate of 3.91%, while a 2-year at 4.76% sitting 88 above implies 3.88% ^{[4][7][10][11][1]}. Call it 3.9% and rounding. Ninety-five basis points is 3.8 quarter-point increments ^[2], and a 3-year yield averages the policy path rather than its peak, so a path that comes back down inside those three years has to climb past 3.8 hikes first. Wolf Richter, who writes Wolf Street, wrote that "these buyers and sellers in the bond market are now counting on multiple additional rate hikes, on top of the rate hike this week" ^[8]. The spread by itself does not separate expected policy from the compensation for owning three-year paper.

His case for the long end rests on a historical range. During past periods of growth and inflation, he says, the spread between the 2-year and the 10-year spent lots of time between 100 and 250 basis points; on Friday it was 25 ^[15][11]. Hold the 2-year at 4.76%, put that spread back at the bottom of the range, and the 10-year is 5.76%; at the top of it, 7.26% ^[3]. The first of those is 75 basis points above Friday's close of 5.01% ^[12][4].

Seventy-five basis points takes roughly 5.8% off the price of a 10-year note bought at par with a 5% annual coupon, whose modified duration is near 7.7 years ^[5]. A year of that coupon is 5.01% ^[12]. So the move implied by the low end of Richter's own spread history costs a Friday buyer more than a year of income ^[7].

The flattening between 4 and 10 years ^[17] may be the market taking the hikes seriously and pricing what they do to growth later, in which case 5.01% is not too low at all. And the 5% line has history on its side: when the 10-year pierced it intraday on October 23, 2023, investors jumped off the fence and the yield fell for two months ^[14]. This time, Richter says, the buying only balanced out the selling and the yield got stuck ^[13].

In my view the front end is the more informative part of the curve this month. The selling has been there: 56 of the 3-year's 144 basis points since the end of February arrived in the three weeks after Warsh's August 28 Jackson Hole speech, 39% of a seven-month move ^[2][3][6]. What would show that wrong is the 2-year sliding back toward the funds rate while the 10-year holds near 5%, because that unwinds the hike pricing without costing duration anything. Richter's own summary is gentler than mine: "So maybe all that happened is that the long-term Treasury yields are in the process of normalizing as the bond market is coming back to life, with the encouragement of the Fed under Warsh" ^[18].

Meanwhile the investor who bought the 3-year on Friday declined 15 basis points to add seven years of maturity ^[5][8]. Richter notes that at 4.86%, within 15 basis points of the 10-year and close to 5%, the demand that level should attract has not emerged, and the 3-year rose another 8 basis points on Friday ^[6].