Front-End Rates and Long-End Relative Value in the U.S. Treasury Curve

Front-end rates provide a conditional reference for relative value between ten-year and thirty-year Treasuries. We evaluate the full link from conditional flatness to spread adjustment, portfolio expression, and supply sensitivity using public U.S. data. In a chronological evaluation beginning in 2015, a lagged reference gap lowers twenty-observation official-curve forecast MSE by 3.48 percent. Removing front-end conditioning reverses that improvement. Frozen-reference accounting associates a one-basis-point initial zero-coupon gap with 0.35 basis points of spread adjustment and 0.31 basis points of gap closure. Under full cash-flow pricing, positive-gap states outperform nonpositive states for the same funded steepener by 1.33 basis points per gross DV01, with positive intervals across four fixed dependence assumptions. The positive-state steepener has a mean net outcome of 0.72 and standard deviation of 3.68, compared with 19.89 for a two-year long. Its lower directional sensitivity provides relative protection during rising-rate episodes. Forecast strength varies across yield conventions and periods; longer dependence blocks and financing wedges reduce the precision of positive absolute means. Known auction quantities leave conditional adjustment essentially unchanged, while prospective supply changes can shift the valuation reference. The results establish front-end-conditioned curve selection and directional duration as distinct portfolio decisions.