1[2 {"id": "C1", "text": "Weekday means of daily 10-year yield changes (2000-2019) are not jointly different at the 5% level (ANOVA F=0.26, p=0.90).", "results_keys": ["anova_f_stat", "anova_p_value", "n_analysed", "conclusion"], "status": "supported", "caveat": "Classical ANOVA assumes i.i.d. equal-variance errors; volatility clustering not addressed."},3 {"id": "C2", "text": "The Monday contrast is +0.05 bp with 95% CI [-0.32, 0.42] bp, i.e. no detectable Monday effect.", "results_keys": ["effect_estimate", "effect_ci_low", "effect_ci_high"], "status": "supported", "caveat": "HC3 robust SE; does not account for serial dependence. Effects below ~0.4 bp cannot be ruled out."},4 {"id": "C3", "text": "Weekday-specific estimation-period means do not improve out-of-sample forecasts over a zero-change baseline (RMSE 4.884 vs 4.876 bp in 2010-2019).", "results_keys": ["baseline_metric", "model_metric", "n_train", "n_test"], "status": "supported", "caveat": "Single split; no formal forecast-comparison test."}5]6