1[2 {"id":"C1","text":"At h=1 the RW-with-drift log-level model has lower RMSE than the naive forecast (0.550 vs 0.997 pp, 44.8% improvement), and the DM test rejects equal accuracy at 5% (p=0.015).","results_keys":["rmse_model_h1","rmse_naive_h1","rel_improvement_rmse_h1","dm_stat_h1","dm_pvalue_h1","conclusion_code_h1"],"status":"supported","caveat":"Only 10 origins; normal-approximation DM without small-sample correction; fragile."},3 {"id":"C2","text":"At h=3 the model has lower RMSE than the naive forecast (1.142 vs 1.420 pp, 19.6%), but the difference is not significant (DM p=0.122).","results_keys":["rmse_model_h3","rmse_naive_h3","rel_improvement_rmse_h3","dm_stat_h3","dm_pvalue_h3","conclusion_code_h3"],"status":"supported","caveat":"Overlapping multi-step errors and noisy long-run variance with n=10."},4 {"id":"C3","text":"Among the baselines, AR(1) beats the naive forecast at both horizons, and the seasonal naive is worst at h=1 but beats the naive at h=3.","results_keys":["rmse_ar1_h1","rmse_ar1_h3","rmse_snaive_h1","rmse_snaive_h3","rmse_naive_h1","rmse_naive_h3"],"status":"supported","caveat":"Point comparisons only; no significance tests among baselines."},5 {"id":"C4","text":"First-release YoY growth of the h=1 targets differs from currently published growth by 0.548 pp on average, with all revisions negative.","results_keys":["mean_abs_revision_h1_targets","revision_h1_targets"],"status":"supported","caveat":"Current vintage includes a benchmark revision; 10 target months."},6 {"id":"C5","text":"RSAFS publication was irregular: the vintages 2025-10-31 to 2026-03-31 end two months before the origin month, while later vintages end one month before.","results_keys":["publication_delays","data_issues"],"status":"supported","caveat":"Month-end granularity of the vintage grid."}7]8