Do retail sales move Treasury yields?
Verdict
- Direction (release_day): Not Significant
- Volatility (release_day): Not Significant
- Volatility (5d): Not Significant
Retail sales barely move the 10-year Treasury — about 1.3× a normal day, just short of statistically significant (p=0.05), with no consistent direction.
Bottom line: Retail sales is a second-tier release for bonds — a hint of extra movement, but not a reliable mover.
The Data
| Dimension | Horizon | Value | Baseline | Test stat | p-value | Verdict |
|---|---|---|---|---|---|---|
| Direction | release_day | -0.8 bps | 0.0 bps | -0.74 | 0.466 | Not Significant |
| Volatility | release_day | 5.7 bps | 4.3 bps | 1.32 | 0.052 | Not Significant |
| Volatility | 5d | 8.4 bps | 10.5 bps | 0.81 | 0.883 | Not Significant |
- Direction (release_day): No consistent direction (p=0.47).
- Volatility (release_day): ≈1.32× a normal day but just short of significant (p=0.05) — a borderline, modest mover.
- Volatility (5d): No elevation over a week.
Methodology
- Events (N): 38
- Window: 2023-01-18 → 2025-12-16
- Baseline: Unconditional distribution of k-trading-day changes
- Look-ahead protected: Yes
- Tests: ttest_1samp_vs_zero, bootstrap_vs_baseline
Caveats
- The release-day reaction is borderline (×1.32, p=0.05) — not quite significant.
- Surprise (vs. forecast) conditioning is not applied.
- Historical statistics for informational purposes only, not financial advice. Results may vary with sample, period, and baseline definition.
Source
- 10-Year Treasury Constant Maturity, Federal Reserve Board via FRED (Tier A) — DGS10
- Advance Monthly Retail Sales release dates, U.S. Census Bureau (via FRED) (Tier A) — Advance Monthly Retail Sales