Reverse Engineering the Yield Curve

Working Paper: NBER ID: w4676

Authors: David K. Backus; Stanley E. Zin

Abstract: Prices of riskfree bonds in any arbitrage-free environment are governed by a pricing kernel: given a kernel, we can compute prices of bonds of any maturity we like. We use observed prices of multi-period bonds to estimate, in a log-linear theoretical setting, the pricing kernel that gave rise to them. The high-order dynamics of our estimated kernel help to explain why first-order, one-factor models of the term structure have had difficulty reconciling the shape of the yield curve with the persistence of the short rate. We use the estimated kernel to provide a new perspective on Hansen-Jagannathan bounds, the price of risk, and the pricing of bond options and futures.

Keywords: Yield Curve; Pricing Kernel; Bond Pricing; Asset Pricing

JEL Codes: G12; E43


Causal Claims Network Graph

Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.


Causal Claims

CauseEffect
upward slope of the yield curve (E43)pricing kernel (D49)
positive autocorrelation of interest rates (E43)pricing kernel (D49)
pricing kernel (D49)bond pricing dynamics (G12)
pricing kernel (D49)option pricing approaches (G13)
mean yield curve (E43)mean reversion in the kernel (C22)
mean yield curve (E43)mean reversion in the short rate (E43)
positively autocorrelated short rate (C22)futures prices lower than forward prices (G13)

Back to index