# Contract (P0)

Question. At the end of each month t (48 origins 2022-05..2026-04), how accurately can Canadian monthly real GDP y/y growth for month t+h (h=1,3) be forecast using only data published by then; does a model adding LFS employment y/y growth beat the naive forecast? Estimands. Out-of-sample RMSE and MAE (pp) per horizon of (a) naive forecast = gdp_yoy(t-2), (b) one model; relative improvement 1 - RMSE_model/RMSE_naive. Meaningful as pseudo-out-of-sample errors on the current vintage (timing simulation only; not real-time / first-release accuracy). 48 origins is a small sample, dominated by the 2022-2026 period. Information rules. At origin t: gdp_* usable only for months <= t-2; emp_* only for months <= t-1. Estimation sample, selection, transformations, plugged-in values and training targets must respect this. Outputs. out/forecasts.csv (192 rows; origin_date,target_date,horizon,forecast,actual,model,last_target_date_used,last_predictor_date_used); out/results.json (keys per TASK.md table); out/report.md (6 sections); out/analysis.py (deterministic regen); out/claims.json. Invalidating conditions. Any use of gdp at months > t-2 or emp > t-1 at origin t; missing actuals for targets (last target 2026-07 is published, so OK); non-reproducible script.