Data: gold/srw_stocks_residual_predictor_analysis.parquet.
Tests variables that may explain the current SRW River Sep stocks residual: production, previous-quarter stocks, exports, lagged model errors, and yield or production deviations in SRW and surrounding states.
Best-practice screen: rank simple one-variable residual adjustments by leave-one-out RMSE improvement, not just in-sample R². The sample is short, so the table favors variables that reduce historical errors when each year is withheld from fitting.
Current use: add the residual adjustment to the baseline SRW River Sep prediction from gold/srw_stocks_prediction.parquet. Positive adjustment means the candidate expects stocks above the baseline model.
Tests the price hypotheses directly using Sep contracts during the July-Aug window. ZW Sep-Dec is Sep Chicago wheat minus Dec Chicago wheat, so more negative means wider carry. ZW-ZC is Sep Chicago wheat minus Sep corn, so lower values mean wheat is cheaper relative to corn.
Leave-one-out residual predictions for the scatter predictor. Values are million bushels.