# Contract (P0)
**Question.** At the end of month t (48 origins 2022-05..2026-04), how accurately can Canadian monthly real GDP yoy growth for month t+h (h=1,3) be forecast using only data published by then, and does a model adding LFS employment yoy growth beat the naive forecast (gdp_yoy[t-2])?
**Estimands.** Out-of-sample RMSE and MAE (pp) over the 48 origins per horizon for naive and one model; relative improvement 1 - RMSE_model/RMSE_naive. Meaningful assuming current-vintage data are an acceptable stand-in (timing simulation only, no real-time vintage claim) and the 48-origin window is representative.
**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 (task_id, model_name, model_description, n_origins, first_origin, last_origin, last_target_date_available, last_predictor_date_available, rmse/mae naive & model h1/h3, rel_improvement_rmse_h1/h3, feasibility, feasibility_reason, data_issues, n_input_rows, n_analysed, random_seed); out/report.md (Question; Data and exclusions; Method; Results; Uncertainty; Limitations and open issues); out/analysis.py; out/claims.json.
**Information rules.** At origin t: GDP columns only for months <= t-2, employment only for months <= t-1; applies to estimation sample, selection, transforms, plug-in values; training targets must be observed at t.
**Invalidity.** Any use of gdp > t-2 or emp > t-1; missing actuals for targets (latest target 2026-07 is available, so feasible); claiming real-time validity.