Does weaker hedging explain the drop in long-term Treasury convenience yields?
Core argument: Applying estimates from a 2005–2024 sample, the authors calculate that the shift in the long-term bond's hedging property can account for approximately a 12 bps drop in Treasury convenience yields in April 2025.
[The convenience yield is proxied by] the yield differential between a synthetic nominal Treasury—constructed out of Treasury Inflation Protected Security (TIPS) and inflation swaps—and a traded nominal Treasury. We construct the TIPS-Treasury premium at the 2, 5, and 10-year horizons using the nominal and real fitted yield curves. Convenience yield erosion for long bonds is consistent with a reduction in the hedging property, reflected in a rising stock-bond covariance. The short end of the Treasury curve, however, continued to exhibit the safe-asset hedging property. April 2025 stands at odds with the historical data in that the covariance between convenience yield innovations and stock prices was positive. Applying estimates from the 2005-2024 sample, this shift in the hedging property of the long-term bond can account for an approximately 12 bps drop in convenience yields in April 2025. These effects are consistent with a withdrawal of safe-asset investors from long-term Treasuries and a rotation towards shorter-term Treasuries and gold. On days where Treasuries were a risk asset - comoving positively with equity returns – gold instead behaved as a hedge, showing negative correlations with equity returns.

